>
volge

Connected and Exposed : State of Play of Operational Technology in the GCC

GCC OPERATIONAL TECHNOLOGY · STATE OF PLAY 2026
The Gulf’s industrial control systems were built to stand alone. Connecting them created a market, and a target.
$4.4B
OT-security market, 2025
~17%
annual growth to 2030
50%
industrial firms hit, 2025
GCC 2026

State of Play is a read on a sector: its size, its structure, where the money moves and what moves it. Each edition takes one market and sets out what is known, what is claimed, and what the difference between the two is worth to anyone deciding where to commit.

The Gulf’s industrial control systems were built to stand alone. Connecting them created a market, and a target. The operational-technology market of the six Gulf states: the systems that run their energy, water and industry, and the security business now growing around their exposure.

Figures carry their basis and their date. Where published estimates disagree, the range is given and the basis named rather than averaged away, and where a widely quoted number does not survive checking it is corrected here rather than repeated. Every chart plots a figure stated in the text, and nothing is projected beyond what a named source projects.

Contents

01Introduction
  1. 02One fact creates both the market and the risk
  2. 05The air gap is gone
  3. 06What OT is, and why the IT playbook fails
  4. 07A market measured in proxies
  5. 09The convergence, in hard numbers
  6. 10The Gulf’s peculiar concentration
02Deep dive
  1. 13How the air gap actually closed
  2. 14The protocols, trusting by design
  3. 16Segmentation, and the standard beneath the market
  4. 17Anatomy of an attack: Shamoon to Triton
  5. 18From episodes to a sustained campaign
  6. 20The world’s OT incidents, for scale
  7. 21Inside the defenders’ toolkit
03Real-world application
  1. 24Saudi Arabia: the connected landscape
  2. 25SABIC and the petrochemical landscape
  3. 26Localising the defence: Aramco × Dragos
  4. 28The national champion: Cyberani’s OT SOC
  5. 29The UAE: ADNOC’s command centre and AI
  6. 30Utilities OT: DEWA × Fortinet
  7. 32Qatar: the OT regulator’s model
  8. 33Kuwait, Bahrain and Oman
  9. 35The giga-projects: building OT from scratch
  10. 36Off-grid by design: Red Sea and Amaala
  11. 38The smart-city landscape
  12. 39Water, power, and the transition multiplier
  13. 41Manufacturing and Industry 4.0
  14. 42The investment layer
04Future outlook
  1. 45Engine one: the capital that builds the assets
  2. 46Engine two: the rulebook that mandates the spend
  3. 48The binding constraint is people
  4. 49The workforce build
  5. 51The frontier: AI on both sides
  6. 52What could go wrong
  7. 53The next two to three years
  8. 55The read, on a page
  9. 56Key takeaways
  10. 57Volge's expertise in operational technology
  11. •Sources
Operational TechnologyState of Play01 •

One fact creates both the market and the risk

One reversible condition created both the Gulf’s fastest-growing security market and its most serious industrial risk.

The air gap is gone, and its dis­ap­pear­ance cre­ated the Gulf’s fast­est-grow­ing se­cur­ity mar­ket and its most ser­i­ous in­dus­trial risk in a single stroke. For dec­ades the sys­tems that run a re­finery, a de­sal­in­a­tion plant or a power grid were pro­tec­ted by phys­ical isol­a­tion. They were “air-gapped”, de­lib­er­ately cut off from cor­por­ate IT and the in­ter­net, and that isol­a­tion was the se­cur­ity. These are not two de­vel­op­ments. They are one.

In­dus­trial di­git­isa­tion de­liv­ers ef­fi­ciency by mak­ing the plant ob­serv­able and con­trol­lable from out­side it. Every con­nec­tion that re­turns data also ac­cepts in­struc­tion, and each is a path that did not exist when the sys­tem was de­signed. So the mod­ern­isa­tion op­por­tun­ity and the se­cur­ity li­ab­il­ity are the same con­nec­tion, seen twice. Three con­sequences mat­ter:

For a re­gion re­build­ing its eco­nomy on con­nec­ted in­dustry, se­cur­ing the op­er­a­tional layer is the pre­con­di­tion for di­git­isa­tion rather than a cost of

it. The order mat­ters, be­cause most budgets are still writ­ten the other way round.

  • The de­mand is struc­tural, not cyc­lical. A di­git­ising in­dus­trial base, a threat his­tory unique to the re­gion, sov­er­eign cap­ital that builds the as­sets, and reg­u­la­tion that man­dates their de­fence all push the same way. The mar­ket’s dir­ec­tion is un­usu­ally cer­tain.
  • The Gulf’s threat lin­eage is the real en­gine. The world’s first mal­ware built to dis­able plant safety sys­tems was found in Saudi Ar­a­bia. The mar­ket grows on memory as much as fore­cast.
  • The de­cis­ive con­straint is cap­ab­il­ity, not cap­ital. Money is abund­ant and the rules are firm; the scarce input is the en­gin­eer­ing tal­ent that can se­cure in­dus­trial sys­tems. Value will ac­crue to who­ever solves the people prob­lem, not who­ever spends the most.

On 3 Novem­ber 2026 the six Gulf states to­gether ac­coun­ted for 100 unique IPv4 ad­dresses re­spond­ing to a nat­ive in­dus­trial-pro­tocol probe, against a world total of 80,029. That is a share of about one eighth of one per cent (Shad­owserver, 2026). Any claim that the Gulf holds thou­sands of in­ter­net-ex­posed in­dus­trial con­trol sys­tems is a de­nom­in­ator error, cor­rec­ted here rather than re­peated. What sur­vives meas­ure­ment is an ar­gu­ment about com­pos­i­tion rather than volume. The re­gion is over-in­dexed

Operational TechnologyState of Play02 •

on the heavy-pro­cess and grid pro­to­cols, and the ex­pos­ure that mat­ters sits at the IT and OT bound­ary and in the vendor re­mote-ac­cess path, neither of which a pub­lic scan can ob­serve. A pub­lic scan sees the wrong sur­face.

The mar­ket is like­wise smal­ler than its head­line fig­ures imply, and con­sid­er­ably less well meas­ured. World­wide spend on cus­tomer-man­aged OT se­cur­ity plat­forms, the de­nom­in­ator that ac­tu­ally cor­res­ponds to the spe­cial­ist vendors, was put at about US$4.37 bil­lion in 2024 (Frost & Sul­li­van, 2025). The fig­ures in the US$23 bil­lion to US$25 bil­lion range bundle con­sult­ing, man­aged ser­vices and in­dus­trial net­work­ing hard­ware along­side the plat­forms. They are not the same thing. At na­tional level, the Saudi reg­u­lator meas­ures the whole Saudi cy­ber­se­cur­ity mar­ket at SAR 16.3 bil­lion for 2025 and pub­lishes no op­er­a­tional-tech­no­logy break­out in any of its three an­nual edi­tions (Saudi NCA, 2026).

The ab­sence of meas­ure­ment is it­self part of the ar­gu­ment. A mar­ket whose own reg­u­lat­ors and ana­lysts can­not yet sep­ar­ate it from en­ter­prise IT is a mar­ket still being defined. That is why the com­pet­it­ive in­tel­li­gence in this read is drawn from pro­grammes, re­gisters and con­tracts rather than from ana­lyst tables. Nobody pub­lishes the num­ber be­cause nobody agrees what to count.

Meas­ure­ment runs the other way on the di­git­isa­tion side. There the count­ing is care­ful. Rys­tad En­ergy puts the cu­mu­lat­ive value of di­git­al­isa­tion and ar­ti­fi­cial in­tel­li­gence to ex­plor­a­tion and pro­duc­tion com­pan­ies at close to 500 bil­lion dol­lars between 2026 and 2030 (Rys­tad En­ergy, 2026). Those same com­pan­ies spent roughly 25 bil­lion dol­lars on di­gital and ar­ti­fi­cial-in­tel­li­gence pur­chases in the pre­ced­ing year. ADNOC re­por­ted 500 mil­lion dol­lars of ar­ti­fi­cial-in­tel­li­gence-de­rived value as early as 2023. It has since com­mit­ted 1.5 bil­lion dol­lars of di­gital cap­ital ex­pendit­ure against a stated tar­get of 1 bil­lion dol­lars a year in value cre­ated. The con­nec­tion that earns the value is the con­nec­tion that car­ries the risk.

That con­trast is the ar­gu­ment of this chapter in its sharpest avail­able form. The be­ne­fit of con­nect­ing a plant is quan­ti­fied to the decimal place by op­er­at­ors and their ad­visers alike, while the li­ab­il­ity the same con­nec­tion cre­ates is re­por­ted in sur­vey per­cent­ages and global in­cid­ent counts. An in­vest­ment com­mit­tee can there­fore price one side of the trade and not the other. The pre­dict­able con­sequence is that mod­ern­isa­tion gets fun­ded on a busi­ness case and pro­tec­tion on a com­pli­ance case. They are one con­nec­tion, seen twice. Until the se­cur­ity budget is de­rived from the di­git­isa­tion budget rather than from an audit find­ing, the se­quence of the spend­ing will keep in­vert­ing the se­quence of the risk.

Operational TechnologyState of Play03 •

The part in brief

  1. 01The air gap is gone, so the modernisation opportunity and the security liability are one connection seen twice, with three consequences for the region.
  2. 02On 3 November 2026 the six Gulf states answered a native industrial-protocol probe with 100 unique addresses, so the claim of thousands of exposed systems is a denominator error.
  3. 03Against a world total of 80,029 responding addresses the Gulf share is about one eighth of one per cent, which makes composition rather than volume the live question.
  4. 04Worldwide spend on customer-managed operational-technology security platforms was about US$4.37 billion in 2024, while headline figures of US$23 billion to US$25 billion bundle consulting, services and hardware.
  5. 05The Saudi regulator put the whole national cybersecurity market at SAR 16.3 billion for 2025 and has published no operational-technology breakout in three annual editions.
Operational TechnologyState of Play04 •
01

The Problem

Why the most connected industrial base on earth is also among the most exposed.

The air gap is gone

The protection that defined industrial security for forty years quietly disappeared, and most operators never decided to remove it.

The Gulf holds one of the densest con­cen­tra­tions of op­er­a­tional tech­no­logy on earth. The con­trol­lers, sensors and safety sys­tems there run oil and gas, water and power, and in­creas­ingly its factor­ies. For most of that land­scape’s his­tory, se­cur­ity meant sep­ar­a­tion. The net­works were phys­ic­ally isol­ated, and an at­tacker on the in­ter­net simply could not reach them. That single as­sump­tion un­der­wrote the en­tire se­cur­ity model, and it is no longer true.

The dis­sol­u­tion was not a de­cision so much as a side-ef­fect. Each ef­fi­ciency pro­gramme (re­mote mon­it­or­ing to cut site vis­its, pre­dict­ive main­ten­ance to re­duce down­time, a di­gital twin to op­tim­ise a pro­cess) needed a link out of the once-sealed net­work. The links ac­cu­mu­lated. The air gap eroded one con­nec­tion at a time while the se­cur­ity model went on as­sum­ing it was in­tact. That is the de­fin­ing con­di­tion of this mar­ket. An in­dus­trial base en­gin­eered for isol­a­tion now runs con­nec­ted, de­fen­ded by a pos­ture built for a world that no longer ex­ists.

Key takeaways
  • OT se­cur­ity once meant isol­a­tion; di­git­isa­tion has re­moved it, link by link.
The air gap was not removed. It was forgotten: assumed to exist while connection quietly dissolved it.
  • The Gulf’s ex­pos­ure is high be­cause its in­dus­trial base is large, valu­able and fast-di­git­ising.
  • The op­por­tun­ity and the risk are the same con­nec­tion: in­sep­ar­able.

The volume fig­ures in cir­cu­la­tion are wrong, and the com­pos­i­tion is the bet­ter evid­ence any­way. On a daily census of the whole IPv4 ad­dress space, the six Gulf states ac­coun­ted for 100 unique ad­dresses re­spond­ing to an in­dus­trial-pro­tocol probe on 3 Novem­ber 2026, fewer than Greece alone. Over the pre­ced­ing two years the com­bined total has moved within a range of 47 to 129 (Shad­owserver, 2026). Held to a con­stant pro­tocol set, Gulf ex­pos­ure fell from about 129 ad­dresses in Janu­ary 2025 to about 47 in June 2026. The ap­par­ent rise later in 2026 is an arte­fact of new probes being added, not of Gulf op­er­at­ors ex­pos­ing new plant. The sur­face did not change.

Low meas­ured ex­pos­ure is not low risk. The census re­cords only what an­swers a nat­ive in­dus­trial probe on a pub­lic ad­dress, and Gulf na­tional oil, power and water op­er­at­ors over­whelm­ingly run op­er­a­tional tech­no­logy be­hind car­rier net­works, private ac­cess-point names, ad­dress trans­la­tion and leased cir­cuits. What the census can­not see is the part that mat­ters:

Operational TechnologyState of Play05 •
Internet-exposed ICS in the GCC, to June 2026
unique IPv4 addresses
037.575112.515061Nov 2024129Jan 2025101Jun 202577Nov 202554Feb 202647Jun 2026
Source: Shadowserver Foundation, 2026

the bound­ary between the busi­ness net­work and the plant, and the re­mote-ac­cess path a vendor en­gin­eer uses. Gulf ex­pos­ure skews to­wards the heavy-pro­cess and grid pro­to­cols rather than to­wards sheer num­bers, which is a state­ment about what is reach­able, not about how much.

Isol­a­tion has not been re­placed by mass ex­pos­ure to the open in­ter­net. It has been re­placed by con­trolled con­nectiv­ity whose con­trols are largely un­doc­u­mented, un­audited and, on the best field evid­ence avail­able, poorly seg­men­ted. That is the nar­rower claim, and it is the stronger one.

Design fact and meas­ure­ment are not the same thing. Com­ment­ary routinely col­lapses the two. Mod­bus in both its serial and TCP forms, DNP3 in its base form, clas­sic S7comm, BACnet/IP, EtherNet/IP and PROFINET RT all ship without nat­ive au­then­tic­a­tion or trans­port

en­cryp­tion, a prop­erty of their spe­cific­a­tions that re­quires no stat­ist­ical source. What no pub­lisher meas­ures is the share of the in­stalled base ac­tu­ally run­ning the se­cure vari­ants. Those vari­ants exist and are sep­ar­ately named: DNP3 Se­cure Au­then­tic­a­tion under IEEE 1815-2012, the OPC UA se­cur­ity modes, PROFINET Se­cur­ity Classes 1 to 3, and CIP Se­cur­ity. A per­cent­age should not be as­ser­ted where only a design fact can be sourced.

What OT is, and why the IT playbook fails

Operational technology looks like IT and behaves like physics. The difference is why securing it is so hard.

Op­er­a­tional tech­no­logy is the com­put­ing that con­trols phys­ical pro­cesses: pro­gram­mable logic con­trol­lers (PLCs), dis­trib­uted con­trol sys­tems (DCS), and su­per­vis­ory con­trol and data ac­quis­i­tion (SCADA). It also takes in the safety in­stru­mented sys­tems (SIS) whose only job is to shut a pro­cess down be­fore it turns dan­ger­ous. Where in­form­a­tion tech­no­logy pro­tects data, OT pro­tects a pro­cess; where IT op­tim­ises for con­fid­en­ti­al­ity, OT op­tim­ises for avail­ab­il­ity and safety. That in­ver­sion breaks the stand­ard se­cur­ity toolkit.

Operational TechnologyState of Play06 •

The reas­ons are con­crete, and they ex­plain why OT se­cur­ity is its own dis­cip­line rather than IT se­cur­ity poin­ted at a fact­ory:

  • You can­not take it down. A re­finery or de­sal­in­a­tion train runs con­tinu­ously; there is no win­dow to patch, and a re­boot can be a safety event.
  • The kit is old and fra­gile. Con­trol­lers are de­signed to last dec­ades; many in ser­vice are un­patch­able and can crash if merely scanned by a con­ven­tional IT tool.
  • The pro­to­cols are trust­ing. OT speaks Mod­bus, DNP3 and OPC: built for closed net­works, with little au­then­tic­a­tion. On the net­work, the sys­tems do as they are told.
  • The fail­ure mode is phys­ical. A breach is not a leaked re­cord but a tripped tur­bine or a dis­abled safety sys­tem: con­sequences meas­ured in bar­rels, mega­watts and lives.

OT se­cur­ity can­not be bolted on by the IT team. It needs tools that watch pass­ively rather than probe, people who un­der­stand both the con­trol­ler and the threat, and an ar­chi­tec­ture built around the pro­cess it­self. None of that is an IT pro­cure­ment with a dif­fer­ent label.

Of ar­chi­tec­ture as­sess­ments con­duc­ted dur­ing 2025, 81 per cent iden­ti­fied poor seg­ment­a­tion between the busi­ness net­work and the plant, and only 46 per cent found ad­equate op­er­a­tional-tech­no­logy net­work mon­it­or­ing de­ployed. Across all in­cid­ent-re­sponse cases to date, 73 per cent in­volved com­prom­ised vir­tual private net­work or jumphost cre­den­tials (Dra­gos, 2026). These are shares of one vendor’s own en­gage­ments rather than

pop­u­la­tion es­tim­ates, and should be read as such and not as a census of in­dustry. The pat­tern holds re­gard­less. The con­trols that fail are ar­chi­tec­tural and ac­cess-re­lated rather than en­d­point-re­lated.

Al­most everything that goes wrong in op­er­a­tional-tech­no­logy se­cur­ity goes wrong at the bound­ary and in the ac­cess path, be­cause the con­trol­ler can­not be patched and the pro­tocol can­not be made to au­then­tic­ate. The fail­ure is ar­chi­tec­tural be­fore it is tech­nical.

A market measured in proxies

The spend is real and accelerating; the precise size is genuinely uncertain, and the uncertainty is itself a finding.

No ana­lyst pub­lishes a clean, GCC-only OT-se­cur­ity fig­ure. The nearest proxy is the Middle East and Africa OT-se­cur­ity mar­ket, put at roughly $4.4 bil­lion in 2025 and grow­ing about 17% a year to­ward nearly $9.7 bil­lion by 2030 (MarketsandMarkets, 2026). At coun­try level, Saudi Ar­a­bia’s in­dus­trial-cy­ber­se­cur­ity mar­ket is es­tim­ated at about $672 mil­lion in 2024, rising to roughly $1.9 bil­lion by 2032: a 14% CAGR, with oil and gas the largest ver­tical at about 35% (P&S In­tel­li­gence, 2025).

The wider con­text sets the ceil­ing rather than the fig­ure. Gart­ner fore­casts MENA IT spend at about $169 bil­lion in 2026, up nearly 9%, with in­form­a­tion-se­cur­ity spend grow­ing faster still. Against those totals, OT se­cur­ity is

Operational TechnologyState of Play07 •

a small but fast-widen­ing slice, the part of the cyber mar­ket grow­ing quick­est be­cause it star­ted latest. The data gap mat­ters. The mar­ket is young enough that the real com­pet­it­ive in­tel­li­gence comes from track­ing pro­grammes and con­tracts, not ana­lyst tables.

MarketSizeSource
MEA OT security$4.4B (2025) → $9.7B (2030)MarketsandMarkets, 2026
Saudi industrial cyber$672M (2024) → $1.9B (2032)P&S Intelligence, 2025
Global OT security~$23–27B (2025/26)Multiple houses
MENA IT spend (context)~$169B (2026)Gartner, 2025

Fig­ures vary by ana­lyst and geo­graphy. The MEA OT fig­ure is Middle East & Africa, not GCC-only. A clean GCC OT-se­cur­ity mar­ket size does not exist pub­licly [In­sert veri­fied GCC-spe­cific fig­ure here when avail­able].

The proxy is weaker than it first ap­pears, in three re­spects worth stat­ing plainly. The head­line re­gional fig­ure is Middle East and Africa, not the Middle East. The pub­lisher’s own page names South Africa as the fast­est-grow­ing single coun­try in that mar­ket, so the geo­graphy in the slug mis­leads. The same pub­lisher sizes the same re­gion in the same year at about US$3.49 bil­lion when the cat­egory is called in­dus­trial con­trol sys­tem se­cur­ity rather than op­er­a­tional tech­no­logy se­cur­ity. A fifth of the mar­ket ap­pears or dis­ap­pears with the defin­i­tion, be­cause the nar­rower re­port car­ries no ser­vices di­men­sion at all. The wider con­text fig­ure, mean­while, is MENA rather than GCC, and no Gulf sub-total is pub­lished any­where in it.

Saudi cybersecurity market, measured by the regulator
SAR bn
0510152013.3202315.2202416.32025
Source: Saudi National Cybersecurity Authority, 2026

Fore­cast sta­bil­ity is also poorer than the decimal places sug­gest. Gart­ner's es­tim­ate of MENA in­form­a­tion-se­cur­ity spend­ing for 2025 was re­vised from US$3.29 bil­lion in April 2025 to US$3.70 bil­lion in Oc­to­ber 2025. That is a rise of 12.5 per cent in six months for a year already half elapsed, al­most all of it in soft­ware. The widely re­cir­cu­lated US$3.3 bil­lion fig­ure for 2025 is stale by the pub­lisher's own reck­on­ing.

Against those mod­elled num­bers stands one meas­ured series. Saudi Ar­a­bia's Na­tional Cy­ber­se­cur­ity Au­thor­ity has now pub­lished three con­sec­ut­ive an­nual meas­ure­ments of its own na­tional mar­ket, with a stated tax­onomy, a pub­lic and private split and an in­tern­ally co­her­ent growth rate. It is the best ma­ter­ial avail­able on Gulf cyber spend­ing, and it con­tains no op­er­a­tional-tech­no­logy break­out at all. For a book about OT in the Gulf,

Operational TechnologyState of Play08 •

the ab­sence of OT from the re­gion's most rig­or­ous na­tional meas­ure­ment is a find­ing rather than a gap to apo­lo­gise for.

The convergence, in hard numbers

The dissolving air gap is not a thesis but a measured condition, and the Gulf inherits it wholesale.

The data on in­dus­trial ex­pos­ure has settled into agree­ment across the major sur­veys. About 40% of or­gan­isa­tions have OT as­sets in­sec­urely con­nec­ted to the in­ter­net, and 55% run four or more sep­ar­ate re­mote-ac­cess tools into their OT en­vir­on­ments (Claroty, 2025). Half of all in­dus­trial or­gan­isa­tions re­por­ted at least one OT se­cur­ity in­cid­ent in the past year (Fortinet, 2025), and Dra­gos logged 1,693 ransom­ware at­tacks on in­dus­trial firms in 2024, up 87% year on year, with man­u­fac­tur­ing the hard­est-hit sec­tor.

The deeper prob­lem is that or­gan­isa­tions can­not see what they are meant to de­fend. Only about 14% re­port full vis­ib­il­ity of their OT land­scape (Fortinet, 2026); and once ex­ec­ut­ive over­sight ex­posed the gaps, the share of firms claim­ing top-tier OT-se­cur­ity ma­tur­ity col­lapsed from 49% to 17% in a single year. The Gulf is no ex­cep­tion: its land­scape is no more vis­ible

than the global av­er­age, while its as­sets are more con­cen­trated, higher-value, and more act­ively tar­geted.

Quick facts: the exposure
OT assets insecurely internet-connected~40%
Firms with four-plus OT remote-access tools55%
Industrial firms with full OT visibility~14%
Industrial ransomware, 2024 (YoY)+87%
Firms at top OT maturity (2025→26)49%→17%

Sources: Claroty State of CPS Se­cur­ity 2025; Fortinet State of OT 2025/2026; Dra­gos 2025. Fig­ures are global; the Gulf is not sep­ar­ately broken out.

Two num­bers in cir­cu­la­tion are wrong, and both should be cor­rec­ted be­fore they travel fur­ther. Com­ment­ary de­scrib­ing thou­sands of in­ter­net-ex­posed in­dus­trial con­trol sys­tems in the Gulf is a de­nom­in­ator error. The veri­fied total was 100 unique IPv4 ad­dresses on 3 Novem­ber 2026 (Shad­owserver, 2026). The word in­cid­ent is the second prob­lem. Within the best open data­set of at­tacks se­lec­ted pre­cisely be­cause they caused phys­ical con­sequences, only 13 per cent in 2024 dir­ectly touched op­er­a­tional-tech­no­logy auto­ma­tion. Close to nine in ten reached phys­ical op­er­a­tions in­dir­ectly, through de­pend­ence on im­paired busi­ness sys­tems (Wa­ter­fall Se­cur­ity and ICS STRIVE, 2025). An OT in­cid­ent, in com­mon usage, is usu­ally an IT event with op­er­a­tional fal­lout.

Operational TechnologyState of Play09 •
Disclosed attacks with physical consequences, heavy industry worldwide
attacks
020406080722023762024572025
Source: Waterfall Security and ICS STRIVE, 2026

The trend in con­sequen­tial at­tacks has also turned, and the turn is in­con­veni­ent for a straight­for­wardly rising nar­rat­ive. Dis­closed at­tacks caus­ing real-world phys­ical dam­age in heavy in­dustry fell about 25 per cent in 2025, the first fall in the stat­istic in six years. Ransom­ware is still the ma­jor­ity of cases, and na­tion-state and hackt­iv­ist activ­ity nearly doubled without off­set­ting the de­cline (Wa­ter­fall, 2026). The same data­set re­cords no new dis­closed in­cid­ents with phys­ical con­sequences in oil and gas in 2024, and names no Gulf state in its re­gional break­down for that year. The hon­est read­ing is dis­clos­ure be­ha­viour rather than safety. Na­tional oil com­pan­ies dis­close al­most noth­ing.

Vis­ib­il­ity has a meas­ur­able price. Av­er­age dwell time for ransom­ware in op­er­a­tional-tech­no­logy en­vir­on­ments was 42 days in 2025 in­dustry-wide,

against 5 days where the or­gan­isa­tion had com­pre­hens­ive OT vis­ib­il­ity (Dra­gos, 2026). That con­trast is the strongest avail­able ar­gu­ment for the in­vent­ory-first se­quence this read re­com­mends.

The Gulf’s peculiar concentration

Why a region with a fraction of the world’s population holds a disproportionate share of its critical OT.

The Gulf’s ex­pos­ure is a struc­tural con­cen­tra­tion, not av­er­age risk ap­plied to large as­sets. The six states sit on a high share of the world’s hy­dro­car­bon pro­duc­tion and a com­mand­ing share of its de­sal­in­a­tion: the re­gion makes roughly 40% of all the de­sal­in­ated water on earth. Where most eco­nom­ies spread crit­ical in­fra­struc­ture across many sec­tors, the Gulf packs it into a hand­ful of enorm­ous, state-owned op­er­at­ors whose con­trol sys­tems are at once the most valu­able to at­tack and the most con­sequen­tial to lose.

Three fea­tures make the con­cen­tra­tion acute. First, the as­sets are ex­ist­en­tial. For sev­eral states, an ex­ten­ded de­sal­in­a­tion out­age is a sur­vival event, not a budget line. Second, own­er­ship is con­cen­trated. A few na­tional cham­pi­ons run most of the land­scape, so one op­er­ator’s pos­ture shapes

Operational TechnologyState of Play10 •

na­tional re­si­li­ence. Third, the land­scape is di­git­ising faster than al­most any­where, be­cause sov­er­eign cap­ital funds mod­ern­isa­tion at a pace private op­er­at­ors can­not match. High value, high con­sequence and high con­nectiv­ity: the three mul­tiply.

Key takeaways
  • The Gulf con­cen­trates crit­ical OT in a few enorm­ous, state-owned op­er­at­ors.
  • De­sal­in­a­tion makes some OT ex­ist­en­tial: fail­ure is days of water, not dol­lars.
  • Sov­er­eign-fun­ded mod­ern­isa­tion di­git­ises the land­scape faster than se­cur­ity can fol­low.

The con­cen­tra­tion is meas­ur­able, and it is more lop­sided than the re­gional fram­ing sug­gests. Across GPCA mem­ber com­pan­ies, Saudi Ar­a­bia held 74.2 per cent of the GCC's 156.2 mil­lion tonnes of pet­ro­chem­ical and chem­ical ca­pa­city in 2023, against Qatar at 10.6 per cent and the UAE at 8.1 per cent. On name­plate re­fin­ing ca­pa­city at end-2024, Saudi Ar­a­bia stood at 3,291 thou­sand bar­rels a day, the UAE at 1,227 and Qatar at 430. Kuwait rose from 800 thou­sand bar­rels a day in 2021 to 1,451 in 2024 as Al-Zour came on stream (OPEC, 2025). Kuwait's in­crease of about 81 per cent in three years is a newly built con­trol es­tate rather than a leg­acy one. That mat­ters for how it is se­cured.

The de­sal­in­a­tion share de­serves more care than it usu­ally re­ceives. The claim that the Gulf pro­duces roughly 40 per cent of the world’s de­sal­in­ated water car­ries no stated basis or year in the sources that cir­cu­late it.

GCC petrochemical and chemical capacity by country, 2023
% of GCC capacity
Saudi Arabia74.2Qatar10.6UAE8.1Oman4.3Kuwait2Bahrain0.9
Source: GPCA, 2024

The com­pet­ing pub­lished fig­ure, 41.8 per cent of cur­rent op­er­a­tional ca­pa­city, is for the Middle East rather than the GCC, and on a dif­fer­ent basis again. Ca­pa­city it­self is re­por­ted in two unit sys­tems and on four in­com­pat­ible bases (con­trac­ted, in­stalled, avail­able and op­er­a­tional). The con­cen­tra­tion ar­gu­ment does not need the per­cent­age. Saudi Ar­a­bia’s own de­sal­in­ated-water pro­duc­tion ca­pa­city rose from 4.6 mil­lion cubic metres a day in 2016 to 16.0 mil­lion at end-2025.

The struc­tural point stands on primary data without it. Name the basis on every fig­ure rather than lean­ing on a world share that nobody ac­tu­ally pub­lishes.

Operational TechnologyState of Play11 •

The part in brief

  1. 01The Gulf's industrial base was engineered for isolation but now runs connected, while the security model continues to assume an air gap that no longer holds.
  2. 02A native-protocol scan on 3 November 2026 returned 100 unique Gulf addresses, fewer than Greece returned, so the exposure is about what is exposed rather than how much.
  3. 03Assessments during 2025 found poor segmentation between business network and plant in 81 per cent of cases, and 73 per cent of incident responses involved compromised remote-access credentials.
  4. 04No clean GCC figure for operational-technology security exists publicly, and the nearest proxy is a Middle East and Africa market of roughly $4.4 billion in 2025.
  5. 05Nearly three quarters of the region's 156.2 million tonnes of petrochemical capacity sat in Saudi Arabia in 2023, which puts the consequence in very few hands.
Operational TechnologyState of Play12 •
The Problem

The air gap was not removed. It was forgotten: assumed to exist while connection quietly dissolved it.

02

The Mechanics

How the exposure is created, how the attacks work, and how the defence is built.

How the air gap actually closed

The mechanism is a forty-year-old reference architecture quietly collapsing under the weight of connectivity.

In prin­ciple, noth­ing from the in­ter­net reaches a con­trol­ler without cross­ing sev­eral guarded lay­ers. The Purdue En­ter­prise Ref­er­ence Ar­chi­tec­ture has or­gan­ised OT net­works for forty years. It places phys­ical pro­cesses at the bot­tom (Level 0–1), con­trol sys­tems above (Level 2–3) and cor­por­ate IT at the top (Level 4–5), with a strictly po­liced bound­ary in between. In prac­tice, the in­dus­trial in­ter­net of things punched holes through every one of them.

The con­ver­gence hap­pens through iden­ti­fi­able chan­nels, and nam­ing them is the first step to de­fend­ing them:

  • Sensors and IIoT. Tens of thou­sands of con­nec­ted sensors stream data up­ward (Ara­mco’s Khurais smart field alone runs more than 40,000), each a device on a once-isol­ated net­work.
  • Re­mote ac­cess. Vendors and en­gin­eers con­nect in to main­tain equip­ment; the av­er­age OT en­vir­on­ment runs four or more re­mote-ac­cess tools, each a door.
  • Di­gital twins and the cloud. Op­tim­ising a pro­cess in the cloud re­quires a live feed from the plant, bridging OT to in­ter­net-fa­cing sys­tems by design.
  • Flattened net­works. To make the data flow, the care­ful Purdue lay­ers are often col­lapsed, so a foothold in IT can reach the con­trol­lers dir­ectly.

Ex­pos­ure is a prop­erty of the mod­ern plant rather than an ac­ci­dent to be elim­in­ated. Re­build­ing the wall is not on offer, be­cause the busi­ness case for con­nectiv­ity is over­whelm­ing. What is left is to make the con­nec­ted land­scape vis­ible, seg­men­ted and watched.

In­dus­trial Eth­er­net ac­coun­ted for 79 per cent of newly in­stalled fact­ory-auto­ma­tion nodes world­wide in 2026, against 34 per cent when the series began in 2015. Field­bus was down to 14 per cent and wire­less steady at 7 per cent (HMS Net­works, 2026). State the de­nom­in­ator with the fig­ure. These are new nodes in a single year, in fact­ory auto­ma­tion, and the series is a vendor es­tim­ate rather than a census. Field­bus re­mains a far lar­ger share of what is ac­tu­ally run­ning in a twenty-year-old re­finery than of what is being bought, and no pub­lisher meas­ures that in­stalled base. That is the num­ber that would mat­ter.

The dir­ec­tion is non­ethe­less un­am­bigu­ous. Each year's new plant is rout­able by de­fault, so the Purdue bound­ary is no longer breached by ex­cep­tion but ab­sent by design at the point of pur­chase. The same in­dustry sur­vey re­cords 93 per cent of re­spond­ents ex­pect­ing op­er­a­tional-tech­no­logy cy­ber­se­cur­ity to change sub­stan­tially over the next five years, al­though the sample is not pub­lished and the fig­ure should be read as sen­ti­ment rather than meas­ure­ment.

Operational TechnologyState of Play13 •
Industrial Ethernet share of newly installed nodes
% of new nodes
020406080342015712024762025792026
Source: HMS Networks, 2026

The work is ar­chi­tec­tural and con­tinu­ous. Field as­sess­ments con­duc­ted dur­ing 2025 found poor seg­ment­a­tion between the busi­ness net­work and the plant in 81 per cent of cases. A fur­ther 82 per cent of or­gan­isa­tions as­sessed had no clear cri­teria for when an op­er­a­tional an­om­aly should trig­ger a cyber in­vest­ig­a­tion (Dra­gos, 2026). Con­nectiv­ity has out­run the con­trols. It has also out­run the pro­ced­ures for in­ter­pret­ing what the con­trols do re­port.

Exposure is a property of the modern plant rather than an accident to be eliminated.

The protocols, trusting by design

Industrial systems speak languages built for a world without attackers. That inheritance is the deepest vulnerability of all.

The reason a foothold on an OT net­work is so dan­ger­ous lies in the pro­to­cols them­selves. Mod­bus, DNP3, OPC and PROFINET were de­signed in and for closed, trus­ted en­vir­on­ments, where every device on the wire was as­sumed le­git­im­ate. They were op­tim­ised for re­li­ab­il­ity and real-time per­form­ance, not se­cur­ity, and most carry little or no au­then­tic­a­tion: a com­mand that ar­rives in the right format is obeyed. Often there is no way for a con­trol­ler to ask “who sent this?”, only to do as it is told.

This is why OT se­cur­ity in­verts the IT model. In IT, the en­d­point can be hardened, patched and au­then­tic­ated. In OT, the en­d­point is a dec­ades-old con­trol­ler that can­not be changed, speak­ing a pro­tocol that can­not be made to au­then­tic­ate. So de­fence moves to the net­work around the device. The traffic is watched for com­mands that should not exist, and the sys­tems that may issue them at all are con­strained. It is also why simply scan­ning an OT net­work with an IT tool can be dan­ger­ous. A probe these pro­to­cols never ex­pec­ted can crash the very device it was meant to as­sess. The pro­to­cols are the in­her­it­ance that makes ar­chi­tec­ture, not patch­ing, the heart of the dis­cip­line.

Operational TechnologyState of Play14 •
Key takeaways
  • OT pro­to­cols (Mod­bus, DNP3, OPC) were built for trus­ted net­works: little or no au­then­tic­a­tion.
  • The con­trol­ler can­not be hardened, so de­fence moves to the net­work around it.
  • Even scan­ning an OT net­work with IT tools can crash fra­gile devices.

This ar­gu­ment can be taken only so far in num­bers. That these pro­to­cols lack nat­ive au­then­tic­a­tion is a spe­cific­a­tion fact and needs no sur­vey. The share of the in­stalled base, or even of the ex­posed base, ac­tu­ally run­ning a se­cure vari­ant is a dif­fer­ent claim, and no pub­lisher sources it. The se­cure op­tions exist and are named, in­clud­ing DNP3 Se­cure Au­then­tic­a­tion, the OPC UA se­cur­ity modes, PROFINET Se­cur­ity Classes 1 to 3 and CIP Se­cur­ity. Nobody meas­ures their de­ploy­ment. Where a per­cent­age is offered for un­au­thentic­ated pro­tocol pre­val­ence, as­sume an in­fer­ence rather than a meas­ure­ment.

One proxy is avail­able, and it needs a pre­cise cap­tion. Of the Gulf’s ex­posed in­dus­trial en­d­points in late 2026, roughly one in four answered on OPC UA, a pro­tocol that at least has a nat­ive se­cur­ity op­tion. The world fig­ure is about one in twenty-eight (Shad­owserver, 2026). The Gulf’s ex­posed land­scape is there­fore markedly more mod­ern than the world’s. An­swer­ing on OPC UA does not mean the se­cur­ity is switched on. The op­tion may sit un­used. The fig­ure meas­ures pres­ence rather than use.

GCC share of the world's exposed ICS, by protocol
% of world exposed IPs
DNP31.8HART-IP1.2ICCP1.2Tank gauging0.9CODESYS0.8OPC UA0.6All protocols0.1
Source: Shadowserver Foundation, 2026

The com­pos­i­tion of Gulf ex­pos­ure is the more dur­able find­ing, and it reads as a heavy-pro­cess and grid pro­file. The baseline share is 0.125 per cent of the world's ex­posed ad­dresses. Against that, the six states hold 1.75 per cent of ex­posed DNP3, 1.18 per cent of HART-IP and 1.17 per cent of ICCP, the inter-con­trol-centre pro­tocol used between util­ity dis­patch centres. They hold es­sen­tially none of the dis­crete-man­u­fac­tur­ing long tail.

Saudi Ar­a­bia has not man­dated IEC 62443. It has man­dated its own Op­er­a­tional Tech­no­logy Cy­ber­se­cur­ity Con­trols, OTCC-1:2022, a na­tional frame­work of four do­mains and 23 sub­do­mains with its own three-level crit­ic­al­ity model, which the reg­u­lator maps to in­ter­na­tional stand­ards, 62443 among them, in a sep­ar­ate annex. An ex­ist­ing 62443 pro­gramme gives an op­er­ator a head start on the OTCC. It does not sat­isfy it. The

Operational TechnologyState of Play15 •

dis­tinc­tion is not pedantic, be­cause the two are audited by dif­fer­ent people. Qatar is closer to the de­scrip­tion above. Its 2023 re­com­mend­a­tions for elec­tri­city and water are built on 62443 it­self.

Segmentation, and the standard beneath the market

If one control defines OT security, it is segmentation; and one standard, IEC 62443, defines how to do it.

Be­cause the con­trol­lers can­not be hardened, OT se­cur­ity leans dis­pro­por­tion­ately on ar­chi­tec­ture: con­trolling what can talk to what. The dom­in­ant frame­work is IEC 62443, the in­ter­na­tional stand­ard for in­dus­trial auto­ma­tion and con­trol-sys­tem se­cur­ity, which Gulf op­er­at­ors in­creas­ingly map their con­trols to and which un­der­pins Saudi and Qatari OT re­quire­ments alike.

Its cent­ral idea is “zones and con­duits”: di­vide the OT land­scape into se­cur­ity zones of sim­ilar risk, and allow traffic between them only through defined, mon­itored con­duits. Done well, a breach in one zone can­not cas­cade into the next, and a com­prom­ised busi­ness net­work can­not reach the safety sys­tems.

For brown­field op­er­at­ors, im­pos­ing this struc­ture on a dec­ades-old, al­ways-on plant is the hard, ex­pens­ive core of the work: the reason OT

se­cur­ity is a multi-year pro­gramme, not a product. For new build, it can be de­signed in from the start, the single largest ad­vant­age the Gulf’s giga-pro­jects hold over the leg­acy land­scape.

  • Zones and con­duits. Group as­sets by risk; allow only spe­cified, in­spec­ted traffic between groups.
  • Se­cur­ity levels. Gradu­ated levels (SL 1–4) match con­trols to con­sequence: a safety sys­tem is held higher than a data his­tor­ian.
  • De­fence in depth. No single con­trol is trus­ted; seg­ment­a­tion, mon­it­or­ing and ac­cess con­trol layer to­gether.
IEC 62443-4-2 component certificates issued, by year
certificates
051015201220183201962020320216202242023152024202025
Source: ISASecure register, 2026

Cer­ti­fic­a­tion is far thin­ner than the stand­ard’s repu­ta­tion im­plies. The en­tire global pop­u­la­tion of ISASecure cer­ti­fic­ates stood at 159 on 4 Novem­ber 2026: 98 cer­ti­fied com­pon­ents, 51 cer­ti­fied de­vel­op­ment or­gan­isa­tions and

Operational TechnologyState of Play16 •

10 cer­ti­fied sys­tems. All ten cer­ti­fied sys­tems are Se­cur­ity Level 1. No Level 4 cer­ti­fic­ate ex­ists on any re­gister, Level 3 ap­pears on only a hand­ful of com­pon­ent cer­ti­fic­ates, and no Gulf en­tity ap­pears as a cer­ti­fic­ate holder any­where. Cer­ti­fic­a­tion of a plant, rather than a product, did not exist at all until the ISASecure ACSSA site-as­sess­ment scheme rolled out in Feb­ru­ary 2026.

A buyer should read all this scep­tic­ally rather than as re­as­sur­ance. The stand­ard defines four se­cur­ity levels and the mar­ket has cer­ti­fied al­most ex­clus­ively to the low­est. The Gulf con­sumes cer­ti­fied equip­ment and pro­duces none of it. Any claim of a 62443-cer­ti­fied Gulf site pred­at­ing 2026 refers to some­thing other than a site cer­ti­fic­ate.

Anatomy of an attack: Shamoon to Triton

The Gulf did not learn OT risk from a report. It learned it from two attacks the rest of the world now studies.

In Au­gust 2012, the Shamoon “wiper” struck Saudi Ara­mco, over­writ­ing roughly 30,000 work­sta­tions and for­cing the com­pany off­line for days. Ara­mco says its hy­dro­car­bon con­trol net­works were isol­ated and un­af­fected, so Shamoon was strictly an IT event, but it made in­dus­trial cyber a board-level con­cern across the Gulf and proved that de­struc­tion, not

theft, could be the goal. A second wave re­turned in 2016–17, hit­ting mul­tiple Saudi en­tit­ies.

The more con­sequen­tial event came in 2017, when mal­ware later named Tri­ton (or Trisis) was found at a Saudi pet­ro­chem­ical plant, widely re­por­ted as Petro Rabigh. Its sig­ni­fic­ance is tech­nical and un­pre­ced­en­ted:

  • It tar­geted safety, not pro­duc­tion. Tri­ton was the first mal­ware ever built to ma­nip­u­late safety in­stru­mented sys­tems: the con­trol­lers that exist solely to shut a plant down be­fore an ex­plo­sion.
  • It aimed at phys­ical con­sequence. By dis­abling the last line of de­fence, it sought the con­di­tions for a cata­strophic phys­ical event, not mere dis­rup­tion.
  • It was a tool, not a one-off. In­vest­ig­at­ors at­trib­uted it to a Rus­sian re­search in­sti­tute; the se­cur­ity com­munity tracks its op­er­at­ors as Xeno­time, treat­ing it as a cap­ab­il­ity that can re­turn.

Between them, Shamoon and Tri­ton did for Gulf OT se­cur­ity what no vendor could: they made the threat con­crete, sov­er­eign and local, and they ex­plain why the re­gion’s spend is driven as much by memory as by fore­cast.

The Tri­ton timeline is bet­ter doc­u­mented than any other Gulf in­dus­trial in­cid­ent, be­cause a United States De­part­ment of En­ergy-fun­ded case study re­con­struc­ted it. Ini­tial ac­cess to the busi­ness net­work was in place by May 2017, at least ninety days be­fore the at­tack proper. On 2 June 2017 the im­planted mal­ware ac­ci­dent­ally shut down a single safety in­stru­mented sys­tem con­trol­ler, halt­ing the re­finery for seven days. En­gin­eers

Operational TechnologyState of Play17 •

at­trib­uted the stop­page to mech­an­ical error. Ac­cess then con­tin­ued un­detec­ted for a fur­ther two months while the ad­versar­ies re­peatedly tested ex­ecut­ables. On 4 Au­gust 2017 the mal­ware was triggered de­lib­er­ately, caus­ing six Triconex con­trol­lers to ini­ti­ate a safety shut­down and a ten-day out­age. The fre­quently cited date of Decem­ber 2017 is the date of pub­lic dis­clos­ure, not of the at­tack.

There was no phys­ical dam­age, and the veri­fied rev­enue loss was ap­prox­im­ately US$938,000, which is small: the case mat­ters for what nearly happened to the safety sys­tem rather than for what it cost, and com­ment­ary im­ply­ing large losses is un­sourced. The same study iden­ti­fied twenty at­tacker tech­niques, sev­en­teen of them pre­curs­ors to the trig­ger­ing event, car­ry­ing forty-one ob­serv­ables, of which four­teen were as­sessed as likely to have been per­cept­ible in the pre­ced­ing ninety days. The na­tional labor­at­ory’s own coun­ter­fac­tual is that com­pre­hen­sion time could have been cut to under ten days.

Shamoon de­serves to be cited for what it was, which is an en­ter­prise wiper. Ara­mco’s own po­s­i­tion is that ex­plor­a­tion, pro­duc­tion and fin­an­cial sys­tems sat on isol­ated net­works and were not breached. Using it as evid­ence of op­er­a­tional-tech­no­logy com­prom­ise is a cat­egory error, how­ever often it is re­peated.

The sharpest il­lus­tra­tion came in early 2026. Dur­ing a wider re­gional con­front­a­tion, Ir­a­nian drones struck OT-rel­ev­ant Gulf en­ergy in­fra­struc­ture, re­por­ted by Qatar’s de­fence min­istry to in­clude a water tank at a Me­saieed

power fa­cil­ity and an in­stall­a­tion at Ras Laf­fan. One cau­tion for buy­ers: much hackt­iv­ist activ­ity is claimed rather than con­firmed, and bey­ond Shamoon and Tri­ton there is no in­de­pend­ently veri­fied case of OT-phys­ical dis­rup­tion at a named Gulf op­er­ator. The dur­able sig­nal is the shift ana­lysts de­scribe (from epis­odic at­tacks to a stand­ing cam­paign), which moves OT de­fence from in­sur­ance against a rare event to a con­tinu­ous op­er­a­tional re­quire­ment.

From episodes to a sustained campaign

What was once a string of landmark incidents is now a continuous, multi-actor campaign against Gulf critical infrastructure.

The threat did not end with Tri­ton; it in­dus­tri­al­ised. Three cur­rents now define it. State-aligned activ­ity per­sists. The IRGC-linked CyberAv3ngers group, tracked in OT circles as Baux­ite, has since late 2023 tar­geted pro­gram­mable logic con­trol­lers across en­ergy, water and chem­ical op­er­at­ors, with con­firmed Middle East vic­tims. Iran’s Peach Sand­storm (APT33) de­ployed cus­tom mal­ware against oil-and-gas and gov­ern­ment tar­gets in the UAE in 2024. Crim­inal ransom­ware has scaled in par­al­lel, man­u­fac­tur­ing the most-hit sec­tor world­wide.

Operational TechnologyState of Play18 •
Key takeaways
  • Three threats run in par­al­lel: state act­ors, ransom­ware crews, hackt­iv­ists.
  • Keep the claimed-vs-con­firmed dis­tinc­tion: most hackt­iv­ist OT claims are un­veri­fied.
  • The dur­able sig­nal is a shift to a sus­tained cam­paign: de­fence is now con­tinu­ous.

The in­dus­tri­al­isa­tion is quan­ti­fied on the crim­inal side and named on the state side. Ransom­ware groups ob­served tar­get­ing in­dus­trial or­gan­isa­tions rose from 50 in 2023 to 80 in 2024 and 119 in 2025. Some 3,300 in­dus­trial or­gan­isa­tions were im­pacted in 2025, and man­u­fac­tur­ing ac­coun­ted for more than two thirds of vic­tims (Dra­gos, 2026). Two fig­ures from that work are routinely con­flated and should be kept apart. The rise of 49 per cent is in the num­ber of groups, while the rise of 64 per cent is in at­tack volume on the pub­lisher's own basis. On the state side, 26 op­er­a­tional-tech­no­logy threat groups are tracked world­wide, of which 11 were act­ive in 2025 and three were newly named.

The Gulf ap­pears in that re­cord by name ex­actly once. One newly named group, tracked as PYROXENE, tar­gets avi­ation, aerospace, de­fence and mari­time op­er­at­ors across the United States, West­ern Europe, Is­rael and the United Arab Emir­ates. It moves from busi­ness sys­tems into op­er­a­tional tech­no­logy through sup­ply-chain com­prom­ise and so­cial en­gin­eer­ing. It is as­sessed as sub­stan­tially over­lap­ping activ­ity aligned to the Is­lamic Re­volu­tion­ary Guard Corps Cyber Elec­tronic Com­mand. Both

Ransomware groups observed targeting industrial organisations
groups
037.575112.51505020238020241192025
Source: Dragos, 2026

PYROXENE and Baux­ite de­ployed de­struct­ive wiper mal­ware against crit­ical in­fra­struc­ture dur­ing the re­gional con­flict of June 2025.

What the re­cord does not con­tain is a con­firmed com­prom­ise of a con­trol sys­tem at a named Gulf op­er­ator since Tri­ton. The veri­fied activ­ity is pre-po­s­i­tion­ing, sup­ply-chain ac­cess and wiper de­ploy­ment, which is a stand­ing cam­paign rather than a series of con­sequen­tial events. That dis­tinc­tion should be pre­served, be­cause it is the one most often lost in re­gional com­ment­ary.

The pat­tern is con­sist­ent and in­struct­ive: at­tack­ers in­creas­ingly tar­get the phys­ical pro­cess, often through the IT side, and in­creas­ingly seek per­sist­ent ac­cess rather than a single strike. The Gulf, with its con­cen­trated,

Operational TechnologyState of Play19 •

high-value land­scape, bench­marks against these in­cid­ents rather than dis­miss­ing them as some­body else’s prob­lem.

The world’s OT incidents, for scale

The Gulf’s exposure is part of a global pattern. The landmark cases abroad show what is at stake at home.

OT at­tacks have moved from the­ory to re­cur­ring real­ity world­wide, and the ca­non­ical in­cid­ents define the threat model Gulf op­er­at­ors now plan against:

  • Stuxnet (2010). The first of­fens­ive cyber-weapon, which ma­nip­u­lated PLCs to dam­age Ir­a­nian cent­ri­fuges: the ori­gin of OT war­fare.
  • Ukraine grid (2015–16). BlackEnergy cut power to roughly 225,000 people; In­dus­troyer then auto­mated grid sab­ot­age, the first mal­ware pur­pose-built to at­tack elec­tri­city trans­mis­sion.
  • Co­lo­nial Pipeline (2021). Ransom­ware on the IT side took a 5,500-mile fuel pipeline off­line: proof an OT op­er­a­tion can be hal­ted without touch­ing its con­trol­lers.
  • FrostyGoop (2024). The first mal­ware to weapon­ise the Mod­bus pro­tocol at scale; it left some 600 apart­ments in a Ukrain­ian city without heat­ing for two days in winter.
  • Volt Typhoon (dis­closed 2024). A China-linked actor pre-po­si­tioned in­side US en­ergy, water and com­mu­nic­a­tions net­works: quiet ac­cess, not im­me­di­ate dis­rup­tion.

Scale is bet­ter con­veyed by de­tec­tion time than by case count, and the trend has re­versed. Global me­dian dwell time, the in­ter­val be­fore an in­truder is de­tec­ted, fell from 205 days in 2014 to 10 days in 2023, then rose to 11 days in 2024 and 14 days in 2025 (Man­di­ant, 2026). The rise is a com­pos­i­tion ef­fect rather than a de­teri­or­a­tion in de­tec­tion. Long-dwell es­pi­on­age and North Korean in­form­a­tion-tech­no­logy-worker op­er­a­tions each car­ried a me­dian of 122 days. The head­line going up there­fore says some­thing about who is in­trud­ing, not about how well de­fend­ers are watch­ing. An ex­hibit that omits the ex­plan­a­tion mis­leads.

India is the closest avail­able ana­logue, and it un­der­cuts the ex­pos­ure story. India runs an in­dus­trial base of com­par­able pro­tocol pro­file and holds 700 in­ter­net-ex­posed in­dus­trial ad­dresses against the Gulf’s 100 on the same scan of 3 Novem­ber 2026. Yet not one in­cid­ent in the In­dian pub­lic re­cord car­ries a con­firmed com­prom­ise of a con­trol sys­tem. The Mum­bai out­age of Oc­to­ber 2020 is uni­ver­sally cited as India’s op­er­a­tional-tech­no­logy in­cid­ent. It was found by the Union Gov­ern­ment in March 2021 to have been caused by human error, with the fact-find­ing

Operational TechnologyState of Play20 •
Global median attacker dwell time before detection
days
062.5125187.52502052014162022102023112024142025
Source: Mandiant M-Trends, 2026

com­mit­tee re­cord­ing no con­clus­ive evid­ence of cyber caus­a­tion. The vendor whose re­search is usu­ally in­voked stated ex­pli­citly that it could not link its find­ings to the out­age.

Ca­non­ical in­cid­ents es­tab­lish what is pos­sible, and little more than that. The gap between pre-po­s­i­tion­ing and phys­ical con­sequence re­mains wide, and the re­gions that look most ex­posed on paper turn out to hold the few­est con­firmed con­trol-sys­tem com­prom­ises of all.

Two claims above can­not be sourced, be­cause the un­der­ly­ing data does not exist. No named ana­lyst house pub­lishes a vendor-level per­cent­age mar­ket share for op­er­a­tional-tech­no­logy, in­dus­trial con­trol sys­tem or cyber-phys­ical-sys­tems se­cur­ity. The mar­ket can be ranked po­s­i­tion­ally and sized in ag­greg­ate, but it can­not be shared out by vendor from any

pub­lished source. The fig­ure of roughly 18 per cent for Fortinet in the OT-fire­wall seg­ment is there­fore not trace­able to a pub­lisher. What is veri­fi­able is a po­s­i­tion rather than a share. Fortinet was named over­all leader in West­lands Ad­vis­ory’s IT and OT Net­work Pro­tec­tion Plat­form Nav­ig­ator for 2025, its third con­sec­ut­ive year. The claim that ser­vices ac­count for about 65 per cent of the mar­ket is like­wise un­sourced. The only meas­ured split loc­ated is Saudi Ar­a­bia’s own, at 51 per cent products and 49 per cent ser­vices in 2025 (Saudi NCA, 2026).

Inside the defenders’ toolkit

OT security is not one product but a stack of categories, and the money is moving across it.

Be­cause the con­trol­lers can­not de­fend them­selves, OT se­cur­ity is de­livered as a stack of spe­cial­ist cat­egor­ies built around them. Gart­ner now con­sol­id­ates these under “cyber-phys­ical sys­tems (CPS) pro­tec­tion plat­forms,” and named Claroty, Dra­gos, No­zomi and Armis as Lead­ers in both its 2025 and 2026 as­sess­ments. The cat­egor­ies a Gulf op­er­ator buys, roughly in the order a pro­gramme ad­opts them:

  • Asset dis­cov­ery and vis­ib­il­ity. You can­not de­fend what you can­not see, and only ~14% of firms have full OT vis­ib­il­ity, so in­vent­ory comes first.
Operational TechnologyState of Play21 •
  • Pass­ive net­work mon­it­or­ing. Watch­ing OT traffic for an­om­alies without prob­ing the fra­gile devices.
  • Se­cure re­mote ac­cess (SRA). The fast­est-grow­ing cat­egory, as spend shifts “from de­tec­tion to pro­tec­tion” and op­er­at­ors close the re­mote-ac­cess doors.
  • OT SOC and man­aged de­tec­tion. Round-the-clock mon­it­or­ing by scarce OT ana­lysts, in­creas­ingly bought as a man­aged ser­vice.
  • Uni­direc­tional gate­ways (data di­odes). Hard­ware that lets data leave the OT net­work but phys­ic­ally pre­vents any­thing en­ter­ing: the strongest seg­ment­a­tion con­trol.
  • OT fire­walls and seg­ment­a­tion. En­for­cing the IEC 62443 zones; Fortinet leads the OT-fire­wall seg­ment.

The com­mer­cially de­cis­ive move is the mi­gra­tion of spend from de­tec­tion (know­ing you were at­tacked) to pro­tec­tion (stop­ping it), and es­pe­cially into se­cure re­mote ac­cess. That is a sign the mar­ket is matur­ing from aware­ness to ac­tion. It is also the open­ing through which the Gulf’s man­aged-ser­vice cham­pi­ons are en­ter­ing.

Quick facts: the vendors
  • Gart­ner CPS Lead­ers (2025 & 2026): Claroty, Dra­gos, No­zomi, Armis.
  • OT-fire­wall share leader: Fortinet (~18%); ser­vices are ~65% of the mar­ket.
  • Fast­est-grow­ing cat­egory: se­cure re­mote ac­cess, as spend shifts to pro­tec­tion.
Planned OT security investment, next 12 to 24 months
% of respondents
Asset visibility and inventory54Threat detection and analytics43Vulnerability management41Secure remote access40Incident response37Configuration management32Identity and access management31
Source: SANS Institute, 2025

The vendor field it­self has moved. In Gart­ner's 2026 as­sess­ment the four lead­ers are Claroty, Dra­gos, No­zomi and Armis, with For­escout, Fortinet and Ten­able po­si­tioned as chal­lengers and Cisco, Hon­ey­well, Mi­crosoft, Palo Alto Net­works and TXOne as niche play­ers. In the For­res­ter Wave for op­er­a­tional-tech­no­logy se­cur­ity of the third quarter of 2026, the lead­ers are Palo Alto Net­works, Claroty and Dra­gos, and the ana­lyst's stated con­clu­sion is that asset dis­cov­ery is no longer a dif­fer­en­ti­ator.

De­nom­in­at­ors mat­ter when siz­ing this stack. Spend on cus­tomer-man­aged plat­form products was put at about US$4.37 bil­lion in 2024, while the ser­vices-in­clus­ive siz­ings run between US$23 bil­lion and US$25 bil­lion. Meas­ur­ing plat­form vendors against the lar­ger fig­ure un­der­states their share roughly five­fold, and the two should never ap­pear on one axis.

Operational TechnologyState of Play22 •

The part in brief

  1. 01Four in five new factory-automation nodes installed worldwide in 2026 were Industrial Ethernet, against about a third in 2015, so fresh plant arrives routable.
  2. 02Industrial protocols such as Modbus and classic S7comm ship without native authentication, so defence moves to the network around a controller that cannot itself be patched.
  3. 03Saudi Arabia mandates its own Operational Technology Cybersecurity Controls of 2022 rather than IEC 62443, so an existing 62443 programme gives a head start without satisfying the national requirement.
  4. 04The Triton malware halted a Saudi refinery for seven days in June 2017 and ten days in August 2017, with verified revenue loss of approximately US$938,000 and no physical damage.
  5. 05The count of ransomware groups seen aiming at industrial targets more than doubled between 2023 and 2025, yet no Gulf control-system compromise has been confirmed since Triton.
Operational TechnologyState of Play23 •
The Mechanics

Exposure is a property of the modern plant rather than an accident to be eliminated.

03

The Gulf Landscape

Who runs the operational technology, how they are securing it, and what the contracts reveal.

Saudi Arabia: the connected landscape

No operator on earth has digitised heavy industry faster, or connected more of it, than Saudi Aramco.

Saudi Ar­a­bia an­chors the Gulf OT mar­ket be­cause it an­chors the Gulf’s in­dustry. Ara­mco holds a max­imum sus­tain­able ca­pa­city of 12 mil­lion bar­rels a day and runs a Fourth In­dus­trial Re­volu­tion Cen­ter, un­veiled in 2019, that has since scaled across its op­er­a­tions. The clearest sig­nal of its con­nectiv­ity is its stand­ing in the World Eco­nomic Forum’s Global Light­house Net­work. In Janu­ary 2025 its North Ghawar fa­cil­ity be­came Ara­mco’s fifth Light­house site, mak­ing it the only in­ter­na­tional en­ergy com­pany with more than three. The re­cog­ni­tion is awar­ded pre­cisely for deep di­git­isa­tion, with re­por­ted gains in­clud­ing pro­duc­tion up more than 8% at one site.

That land­scape is both enorm­ous and among the most con­nec­ted in the world. It is the op­por­tun­ity for OT-se­cur­ity pro­viders and the ex­pos­ure for the king­dom in the same set of as­sets, which is why oil and gas takes the largest share of Saudi in­dus­trial-cyber spend. The king­dom’s an­swer is also the most de­veloped in the Gulf, and it has three parts: a bind­ing rule­book, an Ara­mco-led sup­ply-chain man­date, and a de­lib­er­ate build-out of do­mestic pro­viders.

Quick facts: Saudi Aramco
Maximum sustainable capacity12 mmbpd
WEF Global Lighthouse sites5
Sensors, Khurais smart field40,000+
4IRC unveiled2019

The phys­ical plant be­hind that di­git­isa­tion is the thing being de­fen­ded, and it is worth stat­ing in its own units. Saudi Ar­a­bia ran nine re­finer­ies with 3,291 thou­sand bar­rels a day of name­plate ca­pa­city at end-2024. Two of those sites are re­cent ad­di­tions rather than leg­acy plant. Jazan enters the series at nil in 2020 and 400 thou­sand bar­rels a day from 2021 (OPEC, 2025). The gas pro­gramme is adding con­trol sur­face faster still. Ara­mco an­nounced more than US$25 bil­lion of con­tract awards on 30 June 2024, in­clud­ing US$12.4 bil­lion for the second phase of Jafurah and US$8.8 bil­lion for the third phase of the Mas­ter Gas Sys­tem. The lat­ter adds roughly 4,000 kilo­metres of new pipeline and 17 new com­pres­sion trains by 2028.

The pipeline fig­ure is the more use­ful meas­ure of the two. Kilo­metres of pipeline and num­bers of com­pres­sion trains are a dir­ect, primary-source count of new op­er­a­tional tech­no­logy, whereas a cap­ital fig­ure is not. Sched­ules have also moved: Jafurah’s first phase was an­ti­cip­ated in the third quarter of 2025 and pro­duc­tion ac­tu­ally began in Decem­ber 2025, an­nounced in Feb­ru­ary 2026, a slip of about two quar­ters.

Operational TechnologyState of Play24 •
Saudi refining capacity by site, 2024
'000 b/d nameplate
Ras Tanura550Satorp, Jubail460Yasref, Yanbu430Samref, Yanbu400Rabigh400Jazan400Sasref, Jubail305Yanbu220
Source: OPEC Annual Statistical Bulletin, 2025

Ara­mco pub­lishes no cy­ber­se­cur­ity or op­er­a­tional-tech­no­logy budget line of any kind. It dis­closes cy­ber­se­cur­ity risk qual­it­at­ively and states that no at­tack has to date been ma­ter­ial to its fin­an­cial per­form­ance. Any fig­ure cir­cu­lat­ing for Ara­mco’s cyber spend­ing is there­fore not a com­pany dis­clos­ure, and nobody should be present­ing it as one.

SABIC has run a cor­por­ate di­git­al­isa­tion pro­gramme since 2019, thread­ing ana­lyt­ics and re­mote mon­it­or­ing through plants de­signed dec­ades ago. That is the brown­field con­ver­gence chal­lenge in its purest form. So the pet­ro­chem­ical land­scape con­cen­trates the mar­ket’s de­fin­ing ten­sion. The highest safety stakes, the old­est in­stalled base and an act­ive di­git­isa­tion drive ar­rive all at once. It is no co­in­cid­ence that the re­gion’s land­mark

at­tack landed in a pet­ro­chem­ical plant, nor that chem­ic­als sit along­side oil and gas at the centre of Gulf OT-se­cur­ity de­mand.

SABIC and the petrochemical landscape

Downstream of the wells sits a second vast OT landscape, chemicals, where the safety stakes are highest.

Pet­ro­chem­ic­als are the Gulf’s second great OT con­cen­tra­tion, and SABIC is its an­chor. Ma­jor­ity-owned by Ara­mco, SABIC turns over roughly $37 bil­lion (2024) across some sixty man­u­fac­tur­ing sites and tens of mil­lions of tonnes of out­put, run by the dis­trib­uted con­trol sys­tems and safety in­stru­mented sys­tems Tri­ton was built to tar­get. From a se­cur­ity stand­point, a pet­ro­chem­ical plant is the highest-con­sequence en­vir­on­ment in the re­gion: its pro­cesses run at tem­per­at­ures and pres­sures where a dis­abled safety sys­tem is not a dis­rup­tion but a dis­aster.

Key takeaways
  • Pet­ro­chem­ic­als are the highest-con­sequence OT en­vir­on­ment: safety sys­tems are the prize.
  • SABIC’s ~60 sites pair old in­stalled base with an act­ive di­git­isa­tion pro­gramme.
Operational TechnologyState of Play25 •
  • The re­gion’s land­mark at­tack (Tri­ton) hit a pet­ro­chem­ical plant: the spend re­flects it.

The age pro­file of the in­stalled base is what makes the brown­field ar­gu­ment con­crete. GCC pet­ro­chem­ical and chem­ical ca­pa­city rose from 94.7 mil­lion tonnes in 2012 to 156.2 mil­lion tonnes in 2023, an in­crease of about 65 per cent in el­even years (GPCA, 2024). Most of that plant was com­mis­sioned within the life­time of a single gen­er­a­tion of con­trol sys­tems. The re­gional in­stalled base is there­fore too new to write off. It is also too old to patch. Within the 2023 totals, ethyl­ene ca­pa­city stood at 19.9 mil­lion tonnes, urea at 15.2 mil­lion and am­mo­nia at 13.4 mil­lion. Saudi Ar­a­bia held 18.9 mil­lion tonnes of the GCC’s 29.3 mil­lion tonnes of poly­mer ca­pa­city.

GCC petrochemical and chemical capacity
million tonnes
05010015020094.72012119.520151312018154.12021156.22023
Source: GPCA, 2024

These fig­ures cover GPCA mem­ber com­pan­ies only, so they are floors rather than na­tional totals. A com­pet­ing ag­greg­ate of around 118 mil­lion tonnes per annum cir­cu­lates for GCC ca­pa­city with no stated basis. Set it aside in fa­vour of the in­dustry body’s own meas­ure­ment. Kuwait’s neg­at­ive ca­pa­city growth over the dec­ade re­flects the spin-off of its agri-nu­tri­ents pro­duc­tion in 2018, a re­port­ing-bound­ary change rather than plant clos­ure, and should not be charted as de-in­dus­tri­al­isa­tion.

The se­cur­ity con­clu­sion is bet­ter sup­por­ted by that growth than by any sur­vey. The ca­pa­city base has grown by two thirds in a dec­ade, in the sec­tor where a dis­abled safety sys­tem is a dis­aster rather than a dis­rup­tion. That is the clearest case in the Gulf for design­ing se­cur­ity into new trains rather than ret­ro­fit­ting it later.

Localising the defence: Aramco × Dragos

The Gulf’s signature OT-security partnership shows the regional pattern in miniature: buy the capability, then build it at home.

In Novem­ber 2023, Ara­mco and the OT-se­cur­ity spe­cial­ist Dra­gos signed a memor­andum of un­der­stand­ing to pro­tect crit­ical in­dus­trial in­fra­struc­ture in the king­dom. The agree­ment is not­able less for the plat­form than for its struc­ture, which cap­tures how the Gulf means to source

Operational TechnologyState of Play26 •

OT se­cur­ity: as a loc­al­ised in­dustry rather than an im­por­ted product. Two build-outs were named, a local hard­ware-as­sembly fa­cil­ity and an OT-cy­ber­se­cur­ity train­ing academy, and Dra­gos opened a Riy­adh of­fice with a re­gional dir­ector.

The deal also showed the flow of cap­ital. Ara­mco Ven­tures is a named in­vestor in Dra­gos, so the op­er­ator is at once a cus­tomer, a trainer of the work­force, and a share­holder in the sup­plier. The ar­range­ment is a tem­plate for the mar­ket. Loc­al­isa­tion rules push de­liv­ery in-coun­try, and sov­er­eign-ad­ja­cent cap­ital funds the pro­viders that can do it. One cau­tion is needed. While the part­ner­ship is firmly an­nounced, pub­lic sources do not con­firm the academy or as­sembly fa­cil­ity are op­er­a­tional, so read it as an an­nounced pro­gramme whose ex­e­cu­tion is still matur­ing [In­sert veri­fied op­er­a­tional status here when con­firmed].

Key takeaways
  • The model is buy-then-loc­al­ise: im­port the cap­ab­il­ity, build the fa­cil­ity and academy at home.
  • Ara­mco is cus­tomer, trainer and in­vestor (via Ara­mco Ven­tures) at once.
  • An­nounced ≠ op­er­a­tional: treat the academy/fa­cil­ity as in pro­gress.

The tem­plate has since been re­peated across the Gulf, and at greater scale in the United Arab Emir­ates than in Saudi Ar­a­bia. At Make it in the Emir­ates in May 2026 the UAE Cyber Se­cur­ity Coun­cil signed three op­er­a­tional-tech­no­logy agree­ments in a single week. The first, with Dra­gos,

es­tab­lishes an OT Cy­ber­se­cur­ity Centre of Ex­cel­lence provid­ing a live op­er­a­tional en­vir­on­ment for at­tack and de­fence scen­arios. The second, with Siemens, cov­ers a joint in­nov­a­tion centre, loc­ally hos­ted se­cur­ity in­fra­struc­ture and de­ploy­ment of its in­dus­trial se­cur­ity product on UAE cloud in­fra­struc­ture. The third, with Hon­ey­well, cov­ers na­tional cyber-re­si­li­ence pro­grammes for op­er­a­tional-tech­no­logy se­cur­ity and crit­ical-in­fra­struc­ture pro­tec­tion. In Oc­to­ber 2026 Fortinet and Ara­mco signed a memor­andum on in­dus­trial cy­ber­se­cur­ity and op­er­a­tional tech­no­logy, ex­pli­citly ex­plor­at­ory and centred on loc­al­ising se­cur­ity solu­tions and ex­pand­ing local tal­ent rather than on im­ple­ment­a­tion.

The own­er­ship has moved since that agree­ment was signed. Dra­gos is no longer simply a ven­ture-backed spe­cial­ist in which a Gulf op­er­ator holds a stake: Ac­cen­ture holds a ma­jor­ity in­vest­ment in the com­pany as at 2026, and Ac­cen­ture’s sep­ar­ate ac­quis­i­tions of two ad­ja­cent vendors now op­er­ate under Dra­gos lead­er­ship. Ara­mco Ven­tures’ po­s­i­tion should be read against that own­er­ship rather than as the de­fin­ing cap­ital re­la­tion­ship.

The com­mer­cial sig­nal in the pat­tern is con­sist­ent. Every one of these agree­ments pairs a for­eign cap­ab­il­ity with a local fa­cil­ity and a train­ing com­mit­ment, in­creas­ingly with loc­ally hos­ted in­fra­struc­ture as well. The spe­cial­ists are not selling into the Gulf so much as build­ing in it. That is what loc­al­isa­tion rules were de­signed to com­pel, and it is what makes de­liv­ery ca­pa­city, rather than product, the scarce asset.

Operational TechnologyState of Play27 •

The national champion: Cyberani’s OT SOC

If Aramco × Dragos shows the import, Cyberani shows the build: a sovereign-owned provider standing up domestic OT defence.

Cy­ber­ani, a com­pany of Ara­mco Di­gital, is the clearest ex­ample of the Gulf’s third tier of pro­viders, the sov­er­eign-backed na­tional cham­pion. It runs a two-site man­aged se­cur­ity op­er­a­tions centre, in Riy­adh and Dhahran, li­censed by the Na­tional Cy­ber­se­cur­ity Au­thor­ity and cov­er­ing both IT and OT around the clock. It has launched com­mer­cial products from Ara­mco in­tel­lec­tual prop­erty. In 2025 it was named a Leader in IDC’s MarketScape for Middle East man­aged de­tec­tion and re­sponse, a not­able po­s­i­tion for a young, re­gion­ally-built firm.

Cy­ber­ani’s sig­ni­fic­ance is struc­tural. The Gulf does not in­tend to stay a buyer of for­eign OT se­cur­ity; it is fund­ing do­mestic pro­viders to own the re­cur­ring man­aged-ser­vice layer, under loc­al­isa­tion man­dates that re­quire in-coun­try de­liv­ery and Saud­iz­a­tion quotas that re­quire local hir­ing. A 2025 part­ner­ship with KPMG to train Saudi OT spe­cial­ists makes the work­force in­tent ex­pli­cit. For in­vestors, the les­son is that the most de­fens­ible po­s­i­tion in this mar­ket may not be the tool or the global spe­cial­ist, but the sov­er­eign-aligned na­tional cham­pion that sits between them and owns the local re­la­tion­ship.

The reg­u­lator's own re­gister puts this po­s­i­tion on a firmer foot­ing than any vendor claim. The en­tity ap­pears on the Na­tional Cy­ber­se­cur­ity Au­thor­ity's list as Middle East In­form­a­tion Tech­no­logy Solu­tions, Cy­ber­ani by Ara­mco. It holds one of only six Tier 1 man­aged se­cur­ity op­er­a­tions centre li­cences in the King­dom, the tier that per­mits ser­vice to gov­ern­ment en­tit­ies and to private own­ers and op­er­at­ors of crit­ical na­tional in­fra­struc­ture. Twenty fur­ther com­pan­ies hold Tier 2 li­cences, which ex­pressly ex­clude that work. The li­cence is the bar­rier to entry, not the brand. Six firms hold it.

The scale of the Ara­mco re­la­tion­ship is also pub­lished. At LEAP 2024 the com­pany agreed to de­liver in­dus­trial-sys­tems cy­ber­se­cur­ity ser­vices to more than 100 Ara­mco in­dus­trial sites across Saudi Ar­a­bia. The al­li­ance an­nounced with KPMG in Decem­ber 2025 cov­ers the train­ing of 100 Saudi OT spe­cial­ists over three years. It also cov­ers a di­gital twin and sim­u­la­tion labor­at­ory, a women in OT se­cur­ity ini­ti­at­ive and a na­tional OT se­cur­ity con­fer­ence. Rev­enue and head­count are not dis­closed, and Ara­mco re­ports neither Ara­mco Di­gital nor Cy­ber­ani as a seg­ment, so no fin­an­cial scale can be stated.

The Ara­mco and Ray­theon joint-ven­ture de­scrip­tion, still in cir­cu­la­tion, is stale and should be dropped. The com­pany trades as Cy­ber­ani by Ara­mco Di­gital, is lis­ted among Ara­mco Di­gital’s af­fil­i­ates, and its cyber port­fo­lio is sold under the par­ent’s masthead. RTX’s cur­rent share­hold­ing, if any, is not dis­closed by any party.

Operational TechnologyState of Play28 •

The UAE: ADNOC’s command centre and AI

Abu Dhabi’s national oil company built one of the world’s most ambitious digital operations, and an AI layer on top of it.

The UAE is the Gulf’s second pole, and ADNOC its in­dus­trial an­chor. ADNOC pro­duces around 4.85 mil­lion bar­rels a day and tar­gets five mil­lion by 2027. It runs the Pan­or­ama Di­gital Com­mand Centre, which ag­greg­ates data across 14 sub­si­di­ar­ies and 65 sites and has been cred­ited with more than $1 bil­lion in busi­ness value. That fig­ure dates from around 2020 and is best read as dir­ec­tional now. On top of the data sits AIQ, an ar­ti­fi­cial-in­tel­li­gence joint ven­ture (with G42) run­ning more than 200 AI tools across op­er­a­tions and val­ued above $1.4 bil­lion after a 2024 re­struc­tur­ing.

The UAE pairs this op­er­ator-level di­git­isa­tion with a fast-matur­ing na­tional pos­ture. The Cyber Se­cur­ity Coun­cil’s In­form­a­tion As­sur­ance Stand­ard, up­dated to ver­sion 2, adds OT and in­dus­trial-con­trol re­quire­ments; the Na­tional Cy­ber­se­cur­ity Strategy runs to 2031; and in 2026 the Coun­cil launched OT Centres of Ex­cel­lence with both Dra­gos and No­zomi. The pat­tern mir­rors Saudi Ar­a­bia’s (heavy op­er­ator di­git­isa­tion, a strength­en­ing rule­book, part­ner­ships that loc­al­ise the spe­cial­ist cap­ab­il­ity), ex­ecuted through the UAE’s own na­tional cham­pi­ons, Help AG and CPX.

Quick facts: ADNOC
Crude capacity~4.85 mmbpd
Panorama: subsidiaries / sites14 / 65
Panorama value (2020, dated)>$1B
AIQ AI tools in operation200+

Two fig­ures above can­not be veri­fied from any primary source, and a third is simply wrong. The Pan­or­ama Di­gital Com­mand Centre’s cov­er­age of 14 sub­si­di­ar­ies and 65 sites, and the busi­ness value of more than US$1 bil­lion at­trib­uted to it, were not con­firmed against any ADNOC dis­clos­ure in the re­search for this read. The site adnoc.ae serves error pages to auto­mated re­trieval, and no ADNOC an­nual or sus­tain­ab­il­ity re­port car­ries an op­er­a­tional-tech­no­logy or cy­ber­se­cur­ity fig­ure of any kind. The AIQ share­hold­ing given above is su­per­seded. Since com­ple­tion on 7 June 2024 the hold­ing has been Presight 51 per cent and ADNOC 49 per cent. ADNOC re­ceived 4 per cent of Presight from G42 in ex­change for 11 per cent of AIQ, a trans­ac­tion that im­plied an AIQ valu­ation above US$1.4 bil­lion. AIQ dis­closes more than 100 data sci­ent­ists and de­velopers and 21 pat­ents filed as at 2026.

The na­tional pos­ture is bet­ter evid­enced than the op­er­ator pos­ture. The In­form­a­tion As­sur­ance Stand­ard is re­por­ted at ver­sion 2.1 of Novem­ber 2025, car­ry­ing 134 con­trols across 15 fam­il­ies and adding op­er­a­tional-tech­no­logy con­trols for the first time. That count rests on vendor

Operational TechnologyState of Play29 •
UAE refining capacity by site, 2024
'000 b/d nameplate
Ruwais (ADNOC)817Jebel Ali (ENOC)210Fujairah FTL 1120Fujairah FTL 280
Source: OPEC Annual Statistical Bulletin, 2025

com­ment­ary rather than on the stand­ard it­self, and should be ver­sion-la­belled wherever it is used. The 2019 Na­tional Cy­ber­se­cur­ity Strategy is the source of the nine des­ig­nated crit­ical sec­tors usu­ally quoted as cur­rent. It states on its own terms that it was in force for three years from June 2019. No suc­cessor strategy or sec­tor des­ig­na­tion has been veri­fied, and the Crit­ical In­form­a­tion In­fra­struc­ture Pro­tec­tion Policy that would carry one is not re­triev­able.

There is no ADNOC cyber de­fence centre, and the claim should be dropped: re­peated search­ing found no name, loc­a­tion, part­ner list or launch. What ex­ists is a pro­gramme. Under the part­ner­ship signed with OpenAI and G42 on 2 Novem­ber 2026, ADNOC be­came the first en­ergy com­pany to de­ploy OpenAI Fron­tier, with cyber de­fence de­livered through Day­break Blue and more than 40,000 ChatGPT En­ter­prise seats across the work­force. ADNOC also runs a CyberAware AI Hack­a­thon with

an ex­pli­cit op­er­a­tional-tech­no­logy cat­egory, its 2026 final held on 4 and 5 Novem­ber.

Utilities OT: DEWA × Fortinet

The clearest utilities case shows OT security applied where failure is measured in water and power, not data.

Dubai Elec­tri­city and Water Au­thor­ity (DEWA) runs one of the re­gion’s most di­git­ised util­it­ies: a smart grid with more than two mil­lion smart meters and, by mid-2024, full cov­er­age of smart water meters. Every meter and auto­mated sub­sta­tion is a new OT node, and a util­ity can­not tol­er­ate down­time to se­cure them, which makes util­it­ies OT both the most crit­ical and the hard­est to de­fend.

DEWA’s re­sponse shows the toolkit in prac­tice. Through its di­gital arm it runs a se­cur­ity op­er­a­tions centre that bridges OT and IT, and it has opened a cyber-se­cur­ity in­nov­a­tion lab. In Oc­to­ber 2025 it partnered with Fortinet for uni­fied pro­tec­tion across OT, IT, cloud and IoT, anchored on a zero-trust ap­proach and AI-driven threat in­tel­li­gence feed­ing its in­form­a­tion-se­cur­ity centre. DEWA has also been as­so­ci­ated with one of the strongest seg­ment­a­tion con­trols avail­able (a one-way data sys­tem, or data diode, isol­at­ing its power and water OT), though that spe­cific con­trol is best cited with care pending fuller con­firm­a­tion. The case points to the des­tin­a­tion

Operational TechnologyState of Play30 •

for Gulf util­it­ies. It is con­nec­ted for ef­fi­ciency, and de­fen­ded by ar­chi­tec­ture and mon­it­or­ing rather than by isol­a­tion.

Key takeaways
  • Util­it­ies OT is the least patch­able and most con­sequen­tial: fail­ure is loss of water or power.
  • DEWA’s an­swer is an OT/IT-bridging SOC plus a uni­fied zero-trust pro­gramme (Fortinet, 2025).
  • Strongest con­trol is phys­ical seg­ment­a­tion (data di­odes): cite spe­cific de­ploy­ments with care.

The cau­tion above about the data diode ex­tends to the meter counts. Neither the fig­ure of more than two mil­lion smart elec­tri­city meters nor the claim of full smart water meter cov­er­age by mid-2024 was veri­fi­able in the re­search for this read. The one-way data sys­tem isol­at­ing DEWA’s power and water op­er­a­tional tech­no­logy re­mains un­con­firmed by any primary source. The reason is it­self a find­ing. DEWA’s an­nual re­port path re­turned an HTTP error and re­dir­ec­ted to a miss­ing-page hand­ler, and the same was true across the re­gion’s util­it­ies. SPPC, EWEC, Kahramaa, OETC, Nama and the Saudi reg­u­lator all served empty bod­ies, fire­wall re­jec­tions or JavaScript shells. Gulf util­ity dis­clos­ure is markedly thin­ner than the pre­ci­sion of the fig­ures quoted from it im­plies.

What can be meas­ured about Gulf util­ity op­er­a­tions comes from the re­gional sys­tem op­er­ator rather than from any na­tional util­ity. In 2025 the

GCC emergency support events by generation lost, 2025
events
037.575112.5150130100 to 500 MW27501 to 1,000 MW15Above 1,000 MW
Source: GCC Interconnection Authority, 2026

GCC In­ter­con­nec­tion Au­thor­ity re­cor­ded 172 emer­gency sup­port cases in mem­ber net­works, tak­ing the cu­mu­lat­ive total since 2009 to about 3,008. It also pub­lished the sever­ity dis­tri­bu­tion: 130 events in­volving loss of 100 to 500 mega­watts, 27 between 501 and 1,000 mega­watts and 15 above a gigawatt. Sep­ar­ately, sup­port has been provided in 265 events caused by re­new­able-gen­er­a­tion fluc­tu­ation since 2023, of which more than 69 fell in 2025 alone. That is be­fore the bulk of an­nounced Gulf solar ca­pa­city is built.

A util­ity that calls on its neigh­bours fif­teen times a year for losses above a gigawatt is de­fen­ded by ar­chi­tec­ture, mon­it­or­ing and mu­tual sup­port, not by isol­a­tion. That is the des­tin­a­tion, and the evid­ence for it is op­er­a­tional rather than com­mer­cial. DEWA’s own di­gital arm is also build­ing the other

Operational TechnologyState of Play31 •

side of the same land­scape: the Warsan green data centre, above 100 mega­watts planned, had its second phase ap­proved in Janu­ary 2026.

Qatar: the OT regulator’s model

Beyond the two anchors, Qatar runs the Gulf’s most explicit, regulation-led approach to securing operational technology.

Qatar’s OT mar­ket is smal­ler than Saudi Ar­a­bia’s or the UAE’s, but its reg­u­lat­ory model is the most de­veloped. In Oc­to­ber 2023 its Na­tional Cyber Se­cur­ity Agency pub­lished re­com­mend­a­tions for se­cur­ing op­er­a­tional tech­no­logy in elec­tri­city and water on the IEC 62443 stand­ard, im­ple­men­ted with the util­ity Kahramaa and re­quir­ing every in­sti­tu­tion to ap­point a ded­ic­ated OT-se­cur­ity spe­cial­ist. The agency and the util­ity jointly trained an ini­tial co­hort of those spe­cial­ists: a re­cog­ni­tion that the rule means noth­ing without the people to ex­ecute it.

Kahramaa’s smart-meter rol­lout gives the pro­gramme a large con­nec­ted land­scape to gov­ern. There are more than 528,000 smart elec­tri­city meters and 460,000 smart water meters by mid-2025, with elec­tri­city fully con­ver­ted and water tar­geted for 2027. QatarEnergy, mean­while, has been re­cruit­ing senior in­dus­trial-con­trol and OT-se­cur­ity en­gin­eers. The coun­try’s largest op­er­ator is build­ing OT de­fence in-house rather than lean­ing

solely on vendors. Qatar shows a dis­tinct­ive route. Lead with the reg­u­lator and the util­ity, and build the work­force along­side the man­date.

The reg­u­lat­ory ar­chi­tec­ture holds, though three de­tails need cor­rect­ing. The Na­tional Cyber Se­cur­ity Agency was es­tab­lished by Emiri De­cree No. 1 of 2021, and the Na­tional Cyber Se­cur­ity Strategy cov­er­ing 2024 to 2030 was launched on 17 Septem­ber 2024. There is no 2024 Qatari crit­ical-in­fra­struc­ture pro­tec­tion law: the 2024 in­stru­ment is the strategy, not a stat­ute, and no sec­tor des­ig­na­tion trace­able to a primary Qatari in­stru­ment could be loc­ated at all. The lists of seven des­ig­nated sec­tors in cir­cu­la­tion cite no des­ig­nat­ing in­stru­ment and should not be used. Kahramaa’s smart-meter fig­ures were like­wise not veri­fi­able here; what is veri­fied is the util­ity’s con­trac­ted ca­pa­city, 537.73 mil­lion im­per­ial gal­lons a day of water and 10,572 mega­watts of elec­tri­city at end-2024.

The work­force struc­ture holds and the nam­ing does not. Qatar’s gradu­ate pipeline runs through the Na­tional Cyber Se­cur­ity Academy’s Fresh Gradu­ates Train­ing Camp, which launched its twelfth co­hort in Au­gust 2026 along­side the first co­hort of a new Cy­ber­se­cur­ity Man­agers Pro­gramme. The academy pub­lishes a full nine-pro­gramme lad­der across aware­ness, ca­reer pro­gres­sion, in­dustry and lead­er­ship bands, and no en­rol­ment, in­take or gradu­ate count what­so­ever. No op­er­a­tional-tech­no­logy or in­dus­trial-con­trol track ap­pears any­where in it, which sits oddly be­side a reg­u­lator that re­quires every in­sti­tu­tion in elec­tri­city and water to ap­point a ded­ic­ated op­er­a­tional-tech­no­logy se­cur­ity spe­cial­ist.

Operational TechnologyState of Play32 •
Qatar LNG capacity target, before and after the February 2024 revision
million tonnes a year
037.575112.515077Capacity at announcement126Previous 2030 target142Target from 25 Feb 2024
Source: QatarEnergy via Gulf Times, 2024

The scale of the land­scape being gov­erned is firmly sourced, and its dir­ec­tion is up­ward. Qatar’s li­que­fied nat­ural gas tar­get was raised from 126 to 142 mil­lion tonnes a year on 25 Feb­ru­ary 2024 with the ad­di­tion of North Field West, and first cargo from that pro­ject has since slipped to the end of 2031. In the first quarter of 2026, 93 per cent of Qatar’s con­tract awards were gas.

The thin­ness is real but un­evenly dis­trib­uted, and in one re­spect Bahrain is ahead of its lar­ger neigh­bours. Bahrain’s Na­tional Cyber Se­cur­ity Centre, es­tab­lished under De­cree 65 of 2020, des­ig­nates seven crit­ical na­tional in­fra­struc­ture sec­tors. It is the only Gulf reg­u­lator to name Crit­ical In­dustry as a stan­dalone sec­tor, which brings man­u­fac­tur­ing and in­dus­trial pro­cess plant into scope on its own terms rather than as an ad­junct of en­ergy. It

has writ­ten six sec­tor-spe­cific stand­ards, with Gas, Elec­tri­city and Oil shar­ing one stand­ard with Crit­ical In­dustry. It is also the only Gulf reg­u­lator found to pre­scribe a cer­ti­fic­a­tion lad­der for crit­ical-in­fra­struc­ture staff, clas­si­fied across three levels and four se­cur­ity func­tions.

Kuwait, Bahrain and Oman

The smaller three run substantial energy OT on thinner domestic security markets, and the evidence gaps are themselves a finding.

The Gulf’s three smal­ler states each run large hy­dro­car­bon OT land­scapes through na­tional oil com­pan­ies, over­seen by na­tional cyber au­thor­it­ies, but with far less pub­lic evid­ence of ded­ic­ated OT pro­grammes than their big­ger neigh­bours:

  • Oman. The fur­thest along of the three: Pet­ro­leum De­vel­op­ment Oman is de­ploy­ing an AI-powered di­gital twin uni­fy­ing OT and IT, and runs a cyber-work­force pro­gramme with Oman­tel; OQ is des­ig­nated crit­ical in­fra­struc­ture. Oman’s over­all cyber mar­ket is small, around $135 mil­lion (2025).
  • Kuwait. A Na­tional Cy­ber­se­cur­ity Cen­ter (es­tab­lished 2022) over­sees a strategy to 2027, and KNPC has named cyber lead­er­ship, but no quan­ti­fied KOC or KNPC OT frame­work is pub­lic. The mar­ket is roughly $89 mil­lion (2025).
Operational TechnologyState of Play33 •
  • Bahrain. A Na­tional Cyber Se­cur­ity Centre runs a 2025–2028 strategy with crit­ical-in­fra­struc­ture re­si­li­ence as its first pil­lar, but no Bapco-spe­cific OT pro­gramme is pub­licly doc­u­mented: a genu­ine evid­ence gap [In­sert veri­fied Bapco OT pro­gramme here if dis­closed].

The map is con­sist­ent. The depth of an OT-se­cur­ity mar­ket tracks the depth of its in­dustry, and for the smal­ler three the pub­lic re­cord is thin. For an in­vestor, that thin­ness is op­por­tun­ity as much as risk. Large en­ergy OT land­scapes with under-built do­mestic se­cur­ity mar­kets are ex­actly where a loc­al­ised pro­vider can es­tab­lish it­self early.

Bahrain peak electricity load
MW
01,0002,0003,0004,0003,66220203,75120213,70820223,81920233,9002024
Source: Bahrain Electricity and Water Authority, 2025

Oman's in­sti­tu­tional place­ment is com­monly mis­described and the cor­rec­tion mat­ters for dis­clos­ure. Royal De­cree 64/2020, is­sued on 10 June 2020, places the Cyber De­fence Centre under the In­ternal Se­cur­ity Ser­vice

and makes no men­tion of the tech­no­logy min­istry at all. The at­tached sys­tem, where any sec­tor des­ig­na­tion would sit, is ex­pli­citly un­avail­able in Eng­lish, so Oman's scope is un­veri­fied rather than ab­sent. Kuwait is the weak­est-doc­u­mented of the six: no primary doc­u­ment from its reg­u­lator or na­tional centre was ob­tain­able, and the two com­pli­ance cata­logues in cir­cu­la­tion give ir­re­con­cil­able struc­tures for its frame­work. Neither is us­able.

The in­dus­trial base be­hind these thin mar­kets is non­ethe­less sub­stan­tial, and in Bahrain’s case lar­ger than its util­ity stat­ist­ics show. The au­thor­ity re­ports 5,044 mega­watts of in­stalled ca­pa­city, 254 primary sub­sta­tions and 10,497 dis­tri­bu­tion sub­sta­tions across 13,245 kilo­metres of un­der­ground cable, while a share-based in­ter­na­tional es­tim­ate im­plies roughly 8.7 gigawatts na­tion­ally in 2020. The gap of about 3.5 gigawatts is most plaus­ibly cap­tive in­dus­trial gen­er­a­tion, which is pre­cisely where in­dus­trial con­trol sys­tems con­cen­trate and pre­cisely what the na­tional util­ity does not count.

The thinness is real but unevenly distributed, and in one respect Bahrain is ahead of its larger neighbours.

That green­field char­ac­ter is both the ad­vant­age and the risk. Build­ing new means OT can be se­cure by design, seg­men­ted and mon­itored and stand­ards-aligned from the out­set, which leg­acy op­er­at­ors can only

Operational TechnologyState of Play34 •

ap­prox­im­ate. But it also means en­tire in­dus­trial cit­ies will run on con­nec­ted, autonom­ous sys­tems with little manual fall­back. There is a not­able, re­port­able gap. Across the major giga-pro­jects, no ded­ic­ated in­dus­trial OT-se­cur­ity vendor is pub­licly named for any single in­dus­trial asset. The vis­ible pos­ture is largely IT-side cer­ti­fic­a­tion and smart-city tool­ing. Whether se­cure-by-design is being real­ised in the OT layer, or merely in­ten­ded, is one of the most con­sequen­tial open ques­tions in the Gulf mar­ket [In­sert veri­fied giga-pro­ject OT-se­cur­ity de­ploy­ment here when dis­closed].

The giga-projects: building OT from scratch

The Gulf’s giga-projects are a different OT market: not legacy systems to retrofit, but infrastructure designed connected from the first drawing.

Saudi giga-pro­ject con­tract awards rose about 20% in 2025 to roughly $196 bil­lion, with com­bined com­mit­ments well above a tril­lion dol­lars. The most am­bi­tious is NEOM. Its in­dus­trial core, Oxagon, is planned as a fully auto­mated, all-elec­tric port and man­u­fac­tur­ing city. Its AI arm, Tonomus, is build­ing a “cog­nit­ive city” op­er­at­ing sys­tem and tar­get­ing on the order of 200 mil­lion con­nec­ted IoT devices by 2030. A 1.5-gigawatt, net-zero DataVolt AI data-centre cam­pus an­chors a pivot to­ward

com­pute and en­ergy. The scale of con­nectiv­ity is un­pre­ced­en­ted, and so is the con­cen­tra­tion of de­pend­ence on soft­ware-defined in­fra­struc­ture.

Quick facts: the giga-projects
  • Saudi giga-pro­ject awards ~$196B in 2025 (+20% YoY); com­mit­ments >$1T.
  • NEOM/Tonomus tar­gets ~200M IoT devices by 2030; DataVolt 1.5 GW.
  • No named in­dus­trial OT-se­cur­ity vendor dis­closed for giga-pro­ject OT as­sets: a real gap.

The cap­ital fig­ures above run in the wrong dir­ec­tion and should be re­placed. Gulf con­tract awards did not rise in 2025. They fell about 32 per cent, from a re­cord US$313.9 bil­lion in 2024 to US$213.2 bil­lion, a de­cline of just over US$100 bil­lion at­trib­uted prin­cip­ally to the Saudi giga-pro­ject slow­down. The com­mit­ments fig­ure is a dif­fer­ent ob­ject again. The three cuts of the un­der­ly­ing data­base, US$3.2 tril­lion tracked, US$2.0 tril­lion up­com­ing and US$2.14 tril­lion for Saudi Ar­a­bia in­clud­ing work under ex­e­cu­tion, are not com­par­able and must never be summed. Against the tracked Saudi giga-pro­ject pipeline of more than US$900 bil­lion stands the audited po­s­i­tion. The Pub­lic In­vest­ment Fund car­ried all five giga-pro­ject com­pan­ies at SAR 211 bil­lion, about US$56.2 bil­lion, at end-2024, down 12.4 per cent on 2023 and 6 per cent of its as­sets. The ratio of roughly six­teen to one between tracked pipeline and bal­ance sheet is the clearest single il­lus­tra­tion of how these num­bers go wrong.

Operational TechnologyState of Play35 •
NEOM's 2030 population target, as revised
people
0375,000750,0001,125,0001,500,0001,500,000Original ambition300,000Revised, c. 2024100,000Revised, May 2026
Source: Semafor, 2026

NEOM's am­bi­tion has been cut harder than the cap­ital. The 2030 pop­u­la­tion tar­get has moved from 1,500,000 to 300,000 and, in May 2026, to up to 100,000. Fur­ther work on The Line is delayed until at least after 2030 and the towers are to be sub­stan­tially re­designed; Tro­jena re­ceives no new in­vest­ment until after 2030; and the Red Sea coast des­tin­a­tions in­side NEOM are post­poned. The US$500 bil­lion fig­ure is a 2017 an­nounce­ment and has never been re­stated in any of­fi­cial dis­clos­ure.

What sur­vives the descop­ing is the part this read is about. Oxagon pro­ceeds with roughly US$3 bil­lion of fur­ther planned spend. The DataVolt cam­pus there, 1,500 mega­watts with a first phase of 300 mega­watts tar­geted for 2028, is privately owned by Vis­ion In­vest with NEOM as land­lord rather than com­pute buyer. The descop­ing re­moves real

es­tate and leaves the port, the util­it­ies and the in­dus­trial city. It leaves the con­trol sys­tems.

Off-grid by design: Red Sea and Amaala

Two destinations are proving a model the rest of the world is only theorising: critical infrastructure that depends entirely on software.

Red Sea Global’s flag­ship des­tin­a­tion runs on what is re­por­ted as the world’s largest off-grid solar-and-stor­age mi­crogrid, on the order of 400 mega­watts of solar and 1.3 gigawatt-hours of bat­tery stor­age, with no con­nec­tion to the na­tional grid. Along­side it sit roughly 2,500 en­vir­on­mental IoT sensors and a smart-des­tin­a­tion plat­form span­ning dozens of ser­vices. Its sis­ter pro­ject, Amaala, runs on a com­par­able off-grid sys­tem of around 250 mega­watts of solar and 700 mega­watt-hours of stor­age. These are not pi­lots. They are op­er­at­ing des­tin­a­tions whose power and water exist only as long as the con­trol soft­ware does.

The se­cur­ity im­plic­a­tion is stark and novel. A con­ven­tional asset can fall back to the grid; an off-grid mi­crogrid can­not. The bat­tery con­trol­lers, in­vert­ers and en­ergy-man­age­ment soft­ware are not an ef­fi­ciency layer over a phys­ical sys­tem: they are the sys­tem, and their com­prom­ise is not a

Operational TechnologyState of Play36 •

de­grad­a­tion but a black­out. Yet the dis­closed se­cur­ity pos­ture for these pro­jects is in­form­a­tion-se­cur­ity cer­ti­fic­a­tion, with no named OT vendor for the mi­crogrid, stor­age or de­sal­in­a­tion layer. The Gulf is build­ing the most soft­ware-de­pend­ent crit­ical in­fra­struc­ture on earth; whether it is se­cur­ing the op­er­a­tional core at the same stand­ard is, for now, un­proven in pub­lic sources.

The cent­ral claim is sound and one of its sup­port­ing as­ser­tions is not. Neither mi­crogrid has a veri­fied ca­pa­city fig­ure in the re­search for this read. The 400 mega­watts of solar and 1.3 gigawatt-hours of stor­age at­trib­uted to the flag­ship des­tin­a­tion, the com­par­able fig­ures for Amaala and the count of roughly 2,500 en­vir­on­mental sensors were not con­firmed against any primary source. The claim that no named se­cur­ity pro­vider ex­ists for these pro­jects is, how­ever, wrong. Sirar by stc signed a ser­vice agree­ment with Red Sea Global on 26 Decem­ber 2023, and lists Red Sea Global, Di­r­iyah Com­pany and Saudi Ar­a­bia Rail­ways among its col­lab­or­at­ive ven­tures. What is ab­sent is not a pro­vider but a dis­closed scope. No party states that the en­gage­ment reaches the mi­crogrid, the stor­age or the de­sal­in­a­tion layer.

Meas­ured re­gional data makes the case that isol­a­tion is now a li­ab­il­ity rather than a de­fence. Re­new­able in­ter­mit­tency is already draw­ing on mu­tual sup­port across the in­ter­con­nec­ted Gulf grid, in 265 re­cor­ded sup­port events since 2023 and more than 69 in 2025 alone. An off-grid des­tin­a­tion has no equi­val­ent re­course. How new this plant is also bears stat­ing. Gulf

GCC renewable generation capacity by country, end-2022
MW
UAE3,592Qatar824Oman705Saudi Arabia443Kuwait97Bahrain12
Source: IRENA, 2023

re­new­able gen­er­a­tion ca­pa­city rose from 176 mega­watts in 2013 to 5,673 mega­watts at end-2022, still only about 3 per cent of in­stalled ca­pa­city. Wind across all six states amoun­ted to 67 mega­watts, three-quar­ters of it in Oman.

The ex­pos­ure is novel in kind and small in quant­ity, which is the hon­est way to put it. These are not pi­lots, and their con­trol­lers are the sys­tem rather than a layer above it. But the Gulf’s off-grid build sits on a re­new­ables base that barely ex­is­ted a dec­ade ago, run by op­er­at­ors whose se­cur­ity pos­ture is dis­closed only as in­form­a­tion-se­cur­ity cer­ti­fic­a­tion.

The central claim is sound and one of its supporting assertions is not.
Operational TechnologyState of Play37 •

The smart-city landscape

Beyond the flagship two, a wave of smart developments is wiring the building, the substation and the street: each a new OT surface.

A second tier of giga-de­vel­op­ments is di­git­ising the urban en­vir­on­ment it­self. Qid­diya, the en­ter­tain­ment city, is build­ing AI city-op­er­a­tions on Google Cloud. Di­r­iyah, the her­it­age de­vel­op­ment, has en­gaged stc’s Giza Sys­tems as mas­ter sys­tems in­teg­rator for some fifty smart ser­vices with di­gital twins of trans­port, en­ergy and waste. ROSHN, the hous­ing de­veloper, is wir­ing build­ing-man­age­ment sys­tems through a Cisco part­ner­ship. In the UAE, Mas­dar City has run dis­trict cool­ing, solar gen­er­a­tion and autonom­ous trans­port for over a dec­ade. Each blends build­ing, en­ergy and mo­bil­ity OT into a single con­nec­ted fab­ric.

The pat­tern across this tier is con­sist­ent. For OT se­cur­ity, it is wor­ry­ing. The dis­closed activ­ity is heav­ily IT-and-ap­plic­a­tion-side (in­form­a­tion-se­cur­ity cer­ti­fic­a­tions, smart-city plat­forms, the oc­ca­sional bug-bounty pro­gramme), while the op­er­a­tional layer be­neath has little named OT-se­cur­ity pro­vi­sion in pub­lic sources. That layer is the build­ing-man­age­ment sys­tems, the dis­trict-cool­ing con­trol­lers and the sub­sta­tions. Smart cit­ies mul­tiply the num­ber of con­nec­ted con­trol sys­tems an at­tacker can reach, often out­side the heav­ily-reg­u­lated en­ergy peri­meter. The Gulf’s urban di­git­isa­tion is ra­cing ahead of its urban OT se­cur­ity, and clos­ing that gap is a mar­ket in wait­ing.

Key takeaways
  • Smart de­vel­op­ments (Qid­diya, Di­r­iyah, ROSHN, Mas­dar) blend build­ing, en­ergy and mo­bil­ity OT.
  • Dis­closed se­cur­ity is mostly IT/ap­plic­a­tion-side; the OT layer is under-pro­vi­sioned.
  • Urban OT sits out­side the en­ergy peri­meter: a fast-grow­ing, under-served sur­face.

Named se­cur­ity pro­vi­sion at this tier does exist, which qual­i­fies the con­clu­sion above. Sirar by stc holds a stra­tegic agree­ment with Riy­adh Air­ports Com­pany of 28 Novem­ber 2024, and worked on King Ab­dul­lah Fin­an­cial Dis­trict dur­ing 2025 on what its par­ent de­scribes as AI-driven cyber de­fence of smart in­fra­struc­ture. It has fur­ther agree­ments with Di­r­iyah Com­pany and Saudi Ar­a­bia Rail­ways. The Saudi In­form­a­tion Tech­no­logy Com­pany signed its own memor­andum with Di­r­iyah on 22 Au­gust 2023. What is miss­ing is dis­closed op­er­a­tional-tech­no­logy scope. Only one en­gage­ment in the whole Saudi pro­vider re­cord names it: Sirar’s agree­ment of 28 Novem­ber 2024 with the Na­tional Water Com­pany to se­cure its op­er­a­tional-tech­no­logy in­fra­struc­ture.

The evid­ence that urban con­trol sys­tems are the weaker sur­face is in­dir­ect but con­sist­ent. In the second quarter of 2026 build­ing auto­ma­tion car­ried the highest share of blocked worms of any in­dustry in the Middle East, at 2.90 per cent, and the share rose across every in­dustry in the re­gion over the quarter. Worms are the sig­na­ture of un­seg­men­ted, un­man­aged net­works rather than of tar­geted in­tru­sion, and build­ing-man­age­ment

Operational TechnologyState of Play38 •

sys­tems are where the Gulf’s urban di­git­isa­tion is thick­est and its reg­u­la­tion thin­nest.

Reg­u­la­tion is the struc­tural reason, and it can be stated pre­cisely. Of the six Gulf states, two pub­lish a list of des­ig­nated crit­ical sec­tors. Bahrain des­ig­nates Crit­ical In­dustry. The United Arab Emir­ates sep­ar­ates elec­tri­city and water but names no crit­ical-in­dustry sec­tor, and Saudi Ar­a­bia's op­er­a­tional-tech­no­logy con­trols name no sec­tors at all, scope being defined func­tion­ally by fa­cil­ity crit­ic­al­ity. Dis­trict cool­ing, sub­sta­tions and build­ing-man­age­ment sys­tems there­fore fall in­side the peri­meter only where an op­er­ator's own crit­ic­al­ity as­sess­ment puts them there. That de­cision is de­volved to the op­er­ator and pub­lished by nobody.

Water, power, and the transition multiplier

The energy transition is, from an OT standpoint, an attack-surface-expansion programme, and the Gulf is running it at scale.

The Gulf’s util­it­ies are both its most crit­ical OT and its fast­est-grow­ing. The re­gion pro­duces roughly 40% of the world’s de­sal­in­ated water across hun­dreds of plants. Its power grids are di­git­ising fast. Saudi Ar­a­bia’s grid op­er­ator has de­ployed more than ten mil­lion smart meters, and

DEWA more than a mil­lion smart water meters at full cov­er­age. Each device is a new OT node on a dis­persed, al­ways-on net­work.

De­car­bon­isa­tion mul­ti­plies the ef­fect. Saudi Ar­a­bia tar­gets roughly half its gen­er­a­tion from re­new­ables by 2030 and the UAE half its mix clean by 2050 (am­bi­tions in­de­pend­ent track­ers con­sider be­hind sched­ule, but dir­ec­tion­ally enorm­ous). Re­new­able and dis­trib­uted gen­er­a­tion is, by design, more con­nec­ted and more dis­persed than the thermal fleet it re­places:

  • More nodes. Every solar in­verter, wind tur­bine, bat­tery sys­tem and smart meter is a con­trol­ler that did not exist be­fore.
  • More re­mote man­age­ment. Dis­trib­uted as­sets are man­aged re­motely by ne­ces­sity, widen­ing the re­mote-ac­cess sur­face that is already the fast­est-grow­ing risk.
  • More soft­ware on the crit­ical path. In­vert­ers and en­ergy-man­age­ment sys­tems sit between the en­ergy and the grid, so their com­prom­ise has phys­ical con­sequence.

The same trans­ition that de­car­bon­ises the Gulf widens its OT at­tack sur­face. For op­er­at­ors and pro­viders, en­ergy strategy and OT-se­cur­ity strategy are no longer sep­ar­able: de­car­bon­isa­tion is one of the largest single drivers of the ex­pos­ure this read de­scribes.

Three premises above should not be pub­lished in the form given. The share of roughly 40 per cent of the world’s de­sal­in­ated water car­ries no stated basis or year in the sources that cir­cu­late it, and the com­pet­ing pub­lished fig­ure is for a dif­fer­ent geo­graphy on a dif­fer­ent basis. The

Operational TechnologyState of Play39 •

Saudi grid op­er­ator’s de­ploy­ment of more than ten mil­lion smart meters and DEWA’s smart water meter cov­er­age were not veri­fi­able in the re­search for this read. The largest prob­lem is the policy tar­gets. Saudi Ar­a­bia’s tar­get of roughly half its gen­er­a­tion from re­new­ables by 2030 and the UAE’s half-clean-by-2050 am­bi­tion were both re­cor­ded as un­veri­fied pri­ors rather than find­ings. The Saudi 130 gigawatt fig­ure and the na­tional tar­gets for Qatar, Kuwait, Oman and Bahrain fall in the same class. None should be stated as policy without a primary source.

Energy exchanged across the GCC interconnection
GWh
02,0004,0006,0008,0001,95220211,55020221,06920234,83820247,9132025
Source: GCC Interconnection Authority, 2026

The trans­ition ar­gu­ment can non­ethe­less be made from meas­ured data, and the meas­ure­ment is un­flat­ter­ing to the start­ing point. Only around 3 per cent of Gulf in­stalled gen­er­at­ing ca­pa­city was re­new­able, vir­tu­ally all of the rest being gas, with oil products and even crude burned at peak in

Saudi Ar­a­bia and Kuwait. The in­ter­mit­tency cost is already being borne re­gion­ally. The in­ter­con­nec­tion au­thor­ity has provided sup­port in 265 re­new­able-fluc­tu­ation events since 2023, more than 69 of them in 2025. Cross-bor­der flows have mean­while risen steeply. Total en­ergy ex­changed reached 7,913 gigawatt-hours in 2025 against 1,069 in 2023, of which 5,811.6 gigawatt-hours was sched­uled com­mer­cial trade.

Water is where the new con­trol land­scape is largest and the dis­clos­ure poorest. No de­sal­in­a­tion elec­tri­city con­sump­tion fig­ure, in ter­awatt-hours or as a share of na­tional gen­er­a­tion, was ob­tain­able for any of the six states. That is the single largest ana­lyt­ical gap in the re­gional power evid­ence base. The new plant is non­ethe­less vis­ible in con­tracts. Kuwait’s up­grade of the North Kabd sewage treat­ment works was awar­ded in the first quarter of 2026 at US$3.2 bil­lion with a ten-year op­er­a­tion and main­ten­ance period. It is de­signed to pro­cess up to one mil­lion cubic metres of wastewa­ter a day, and is a single large op­er­a­tional-tech­no­logy in­stall­a­tion in its own right.

None of the pro­gramme fig­ures above could be veri­fied, and the fore­cast should carry its pub­lisher’s lim­it­a­tions. The tar­get of 4,000 trans­formed factor­ies and the sub­sidy of up to 80 per cent of di­git­isa­tion cost were not veri­fi­able in the re­search for this read. Neither were the Emir­ati in­dus­trial-GDP pro­gramme fig­ures, nor the smart-man­u­fac­tur­ing fore­cast of US$3.8 bil­lion rising to US$11.9 bil­lion. The last comes from a pub­lisher that is not among the named ana­lyst houses whose scope defin­i­tions were

Operational TechnologyState of Play40 •

checked. The one sourced man­u­fac­tur­ing series loc­ated is a fore­cast of all cy­ber­se­cur­ity spend in that ver­tical, and it car­ries a meth­od­o­lo­gical fault worth nam­ing. The pub­lisher ap­plies a single re­gional sec­tor mix to each coun­try total, so the fig­ures carry in­form­a­tion about scale and noth­ing us­able about sec­tor struc­ture.

Manufacturing and Industry 4.0

The Gulf is building a manufacturing base from a low start, and digitising it as it builds, creating OT that is born connected.

Bey­ond en­ergy and util­it­ies, in­dus­trial di­ver­si­fic­a­tion is cre­at­ing a new man­u­fac­tur­ing OT land­scape. Saudi Ar­a­bia’s Fu­ture Factor­ies pro­gramme tar­gets the trans­form­a­tion of 4,000 factor­ies to In­dustry 4.0 stand­ards, sub­sid­ising up to 80% of the di­git­isa­tion cost. The king­dom’s smart-man­u­fac­tur­ing mar­ket is fore­cast to grow from $3.8 bil­lion in 2025 to $11.9 bil­lion by 2034 (IMARC). The UAE’s Op­er­a­tion 300bn aims to lift in­dus­trial GDP from AED 133 bil­lion to AED 300 bil­lion by 2031 through fourth-in­dus­trial-re­volu­tion ad­op­tion, with man­u­fac­tur­ing grow­ing 7.7% in the first quarter of 2025.

This new OT dif­fers from the leg­acy land­scape in one de­cis­ive re­spect. It can, in prin­ciple, be built se­cure from the start. But man­u­fac­tur­ing is also the most ransom­ware-hit sec­tor on earth, and the pace of the build-out

makes se­cure-by-design the ex­cep­tion rather than the rule. The stra­tegic point is that the Gulf’s in­dus­trial di­ver­si­fic­a­tion and its OT-se­cur­ity mar­ket are the same pro­gramme seen from two angles. Every fact­ory the re­gion adds is both a unit of eco­nomic am­bi­tion and a unit of at­tack sur­face.

Manufacturing is the most ransomware-hit sector on earth. The Gulf is building its factory OT just as it joins the network.
UAE manufacturing cybersecurity spend, forecast
US$ m
062.5125187.5250130.22025145.920261632027181.52028197.120292142030
Source: MarketsandMarkets, 2026

Reg­u­la­tion is the more in­form­at­ive lens, and on it the Gulf treats man­u­fac­tur­ing in­con­sist­ently. Bahrain is the only state of the six to des­ig­nate Crit­ical In­dustry as a stan­dalone crit­ical-in­fra­struc­ture sec­tor, bring­ing in­dus­trial

Operational TechnologyState of Play41 •

pro­cess plant into scope on its own terms. It then writes a single com­bined stand­ard cov­er­ing that sec­tor and Gas, Elec­tri­city and Oil to­gether. Saudi Ar­a­bia's op­er­a­tional-tech­no­logy con­trols name no sec­tors at all. A fact­ory falls in scope only where its own crit­ic­al­ity as­sess­ment, con­duc­ted with the reg­u­lator's fa­cil­ity-level iden­ti­fic­a­tion tool, places it there. The United Arab Emir­ates des­ig­nates no crit­ical-in­dustry sec­tor under the only sec­tor list it has pub­lished.

The stra­tegic point stands, with a sharper edge. In­dus­trial di­ver­si­fic­a­tion and op­er­a­tional-tech­no­logy ex­pos­ure are the same pro­gramme seen from two angles. But the new factor­ies are being added into a reg­u­lat­ory peri­meter that was drawn around hy­dro­car­bons and util­it­ies, and in four of six states the peri­meter’s ex­tent is not pub­lished at all. Se­cure by design is avail­able at the point of pur­chase and com­puls­ory al­most nowhere.

The investment layer

Capital is flowing into OT security from two directions: global venture funding the specialists, and sovereign capital building the regional champions.

The spe­cial­ists are well cap­it­al­ised and con­sol­id­at­ing into a re­cog­nised cat­egory. Claroty raised about $150 mil­lion at a roughly $3 bil­lion valu­ation and No­zomi about $100 mil­lion. Gart­ner now form­al­ises

“cyber-phys­ical sys­tems pro­tec­tion” as a dis­tinct buy­ing mar­ket, a sign OT se­cur­ity has ma­tured from a niche into a fund­able sec­tor. The Gulf adds a stra­tegic, sov­er­eign layer on top. Ara­mco Ven­tures is a named in­vestor in Dra­gos; the na­tional cham­pi­ons (Cy­ber­ani, SITE, Sirar) are fun­ded to build do­mestic ca­pa­city; and the state in­vests dir­ectly. At Black Hat MEA 2025 in Riy­adh, the Saudi Min­istry of In­vest­ment un­veiled more than SR 500 mil­lion (around $133 mil­lion) in new cy­ber­se­cur­ity com­mit­ments.

The in­vest­ment thesis is un­usu­ally clean. It is a struc­tur­ally grow­ing, reg­u­la­tion-backed, re­cur­ring-rev­enue mar­ket, at­tached to the re­gion’s highest-value as­sets, with loc­al­isa­tion rules that pro­tect do­mestic pro­viders. For an in­vestor, the live ques­tion is not whether the mar­ket grows but which layer cap­tures the re­cur­ring spend: the tools, the man­aged ser­vices, or the sov­er­eign-backed na­tional cham­pi­ons that in­creas­ingly sit between the two and own the local re­la­tion­ship.

Quick facts: the capital in
  • Spe­cial­ists fun­ded: Claroty ~$150M (~$3B val), No­zomi ~$100M.
  • Sov­er­eign cap­ital: Ara­mco Ven­tures in Dra­gos; na­tional cham­pi­ons fun­ded to build.
  • Dir­ect state in­vest­ment: ~$133M in new cyber com­mit­ments, Black Hat MEA 2025.

The spe­cial­ists dis­close less than the com­ment­ary as­sumes, so each of the cap­ital fig­ures above needs re­stat­ing. Claroty raised US$150 mil­lion in a Series F on 22 Janu­ary 2026, led by Golub Growth, with con­firmed

Operational TechnologyState of Play42 •

par­ti­cip­a­tion from ex­ist­ing in­vestors of up to a fur­ther US$50 mil­lion. No dol­lar valu­ation was dis­closed. The com­pany pub­lished only an 80 per cent in­crease in valu­ation since an un­dis­closed March 2024 base, so the fig­ure of roughly US$3 bil­lion is not a com­pany state­ment. No­zomi’s round was not veri­fi­able, and its own­er­ship is now an open ques­tion, a re­por­ted ac­quis­i­tion by Mit­subishi Elec­tric at ap­prox­im­ately US$1 bil­lion agreed in Septem­ber 2025 hav­ing no con­firmed com­ple­tion. Dra­gos is no longer a ven­ture-backed in­de­pend­ent, with Ac­cen­ture hold­ing a ma­jor­ity in­vest­ment and runZero and NetRise op­er­at­ing under Dra­gos lead­er­ship, and the valu­ation at­trib­uted to that trans­ac­tion is un­veri­fied.

Sirar by stc segment revenue, audited
SAR m
02505007501,000430.32022588.62023732.72024826.52025
Source: stc Annual Report, 2026

The cat­egory form­al­isa­tion cited above is real, and it should be dated cor­rectly. The Magic Quad­rant for cyber-phys­ical sys­tems pro­tec­tion

plat­forms was first pub­lished on 12 Feb­ru­ary 2025 and re­is­sued on 3 March 2026, su­per­sed­ing two mar­ket guides whose latest edi­tions are 2023 and which are still hos­ted on vendor land­ing pages. On the sov­er­eign side, the only audited fin­an­cial series for a Gulf cyber cham­pion be­longs to Sirar by stc. The fig­ure in cir­cu­la­tion for it is wrong by a factor of about 3.75. The US$827 mil­lion com­monly quoted for 2025 is SAR 826.5 mil­lion, roughly US$220 mil­lion, the dol­lar sign in the par­ent’s chart being a mis-rendered riyal glyph against an audited note in the same doc­u­ment.

The live cor­por­ate ac­tion in the Saudi cham­pion tier is a list­ing rather than a con­sol­id­a­tion. No mer­ger or com­bin­a­tion of the Pub­lic In­vest­ment Fund's cyber as­sets is evid­enced. What is on the re­cord is pre­par­a­tion for an ini­tial pub­lic of­fer­ing of the Saudi In­form­a­tion Tech­no­logy Com­pany, ad­visers ap­poin­ted in Oc­to­ber 2025, with no list­ing com­pleted as at Novem­ber 2026.

Operational TechnologyState of Play43 •

The part in brief

  1. 01Aramco announced more than US$25 billion of awards on 30 June 2024, including US$8.8 billion adding roughly 4,000 kilometres of pipeline and 17 compression trains by 2028.
  2. 02Six firms hold the Saudi managed security operations centre tier that permits service to critical national infrastructure, against twenty in the tier that expressly excludes that work.
  3. 03One engagement in the Saudi provider record names operational-technology scope, a Sirar agreement with the National Water Company signed in late November 2024.
  4. 04NEOM's 2030 population target fell from 1,500,000 to up to 100,000 by May 2026, and what survives the descoping is the port, the utilities and the control systems.
  5. 05Gulf contract awards fell about 32 per cent in 2025, from a record US$313.9 billion to US$213.2 billion, attributed principally to the Saudi giga-project slowdown.
Operational TechnologyState of Play44 •
The Gulf Landscape

Manufacturing is the most ransomware-hit sector on earth. The Gulf is building its factory OT just as it joins the network.

04

The Read

The engines, the constraint, the frontier, and where the market goes by 2028.

Engine one: the capital that builds the assets

OT-security demand in the Gulf is, in effect, a fixed proportion of the largest industrial-investment programme in the world.

The first en­gine of the mar­ket is sov­er­eign cap­ital, and it is vast. Ara­mco’s cap­ital pro­gramme runs above $50 bil­lion a year, with 2026 guid­ance of $50–55 bil­lion. ADNOC has ap­proved a $150 bil­lion plan for 2026–2030. Across the Gulf, na­tional oil com­pan­ies are ex­pec­ted to spend on the order of $115–125 bil­lion a year through 2027. Every bar­rel of that capex builds or con­nects in­dus­trial as­sets that must then be se­cured, so OT-se­cur­ity de­mand grows as a near-mech­an­ical func­tion of in­dus­trial in­vest­ment.

Loc­al­isa­tion turns that de­mand in­ward. Ara­mco’s in-coun­try value pro­gramme has reached about 70% local con­tent, tar­get­ing 75% by 2030, and the UAE’s equi­val­ent ma­ter­i­ally lowers the tender scores of vendors without an in-coun­try pres­ence. Sov­er­eign cap­ital does not merely fund OT se­cur­ity. It re­quires that the se­cur­ity be de­livered within the re­gion, and in­creas­ingly built there. For pro­viders, mar­ket ac­cess is con­di­tional on a local foot­print. That is pre­cisely why the na­tional-cham­pion tier ex­ists, and why it is the most de­fens­ible po­s­i­tion in the mar­ket.

Quick facts: the capital engine
Aramco capex (2026 guidance)$50–55B
ADNOC plan (2026–2030)$150B
GCC NOC capex / year, to 2027~$115–125B
Aramco local-content target (2030)75%

The en­gine is real, and in 2025 it slowed. Gulf con­tract awards fell about 32 per cent, from a re­cord US$313.9 bil­lion in 2024 to US$213.2 bil­lion, and con­tin­ued to soften through the first half of 2026. The coun­try rank­ing also re­versed. In the first quarter of 2026 the United Arab Emir­ates over­took Saudi Ar­a­bia as the largest pro­jects mar­ket in the re­gion, tak­ing 47.7 per cent of Gulf awards against Saudi Ar­a­bia’s 18 per cent. That was Saudi Ar­a­bia’s second-low­est quarter in twenty-one. The sov­er­eign in­vestor it­self shows the same turn, as­sets under man­age­ment fall­ing SAR 38 bil­lion in 2025, its first de­cline of the dec­ade, with the share­holder re­turn mov­ing from 7.2 per cent to 5.8 per cent.

None of the cap­ital-pro­gramme fig­ures quoted above could be veri­fied, and nor could the loc­al­isa­tion per­cent­ages. No primary con­firm­a­tion was ob­tained for a Saudi pro­gramme above US$50 bil­lion a year, for the Emir­ati plan of US$150 bil­lion over 2026 to 2030, or for the re­gional na­tional-oil-com­pany range of US$115 bil­lion to US$125 bil­lion a year. In-coun­try value at about 70 per cent rising to 75 per cent by 2030 is like­wise un­con­firmed. Where a per-pro­ject fig­ure is wanted at all, the sov­er­eign in­vestor

Operational TechnologyState of Play45 •
GCC project contract awards, Q1 2026
US$ bn
UAE29.2Saudi Arabia11Qatar8.8Kuwait8.1Bahrain and Oman4.1
Source: MEED Projects via Kamco Invest, 2026

pub­lishes none. There is no pro­ject-level break­down, no timelines, no pro­jec­ted re­turns and no impair­ments by pro­ject.

The mech­an­ism sur­vives the cor­rec­tion, be­cause de­mand tracks as­sets built rather than dol­lars an­nounced. The bet­ter unit is phys­ical: 4,000 kilo­metres of new pipeline and 17 new com­pres­sion trains by 2028 under one Saudi gas pro­gramme, or one mil­lion cubic metres a day of new Kuwaiti wastewa­ter ca­pa­city. These are counts of con­trol sur­face, and un­like cap­ital fig­ures they do not move when a budget is re­vised.

The struc­ture of the Saudi con­trols is as stated and the scope is not. The op­er­a­tional-tech­no­logy con­trols, is­sued on 2 June 2022 with the Ar­abic text bind­ing, run to 4 do­mains, 23 sub­do­mains, 47 main con­trols and 122 sub­con­trols, and they name no sec­tors what­so­ever. Com­ment­ary list­ing their sec­tors as oil and gas, elec­tri­city and water, pet­ro­chem­ic­als and trans­port is ex­tra­pol­at­ing. Scope is func­tional. The con­trols apply to

in­dus­trial con­trol sys­tems in fa­cil­it­ies deemed crit­ical, in the King­dom or abroad, and crit­ic­al­ity is as­signed by the op­er­ator using the reg­u­lator's own fa­cil­ity-level iden­ti­fic­a­tion tool. That de­vol­u­tion is the struc­tural reason no count of in-scope en­tit­ies ex­ists or can be pub­lished. Com­pli­ance is also ex­pli­citly sub­or­din­ate to op­er­a­tions, re­quired only after the op­er­ator has sat­is­fied it­self that ap­ply­ing a con­trol will not jeop­ard­ise con­tinu­ity.

Engine two: the rulebook that mandates the spend

In the Gulf, OT-security regulation does not merely set a standard: it turns good practice into compulsory, recurring expenditure.

The second en­gine is reg­u­la­tion. Saudi Ar­a­bia’s Na­tional Cy­ber­se­cur­ity Au­thor­ity is­sued ded­ic­ated Op­er­a­tional Tech­no­logy Cy­ber­se­cur­ity Con­trols in 2022. These are bind­ing min­imum con­trols for crit­ical-in­fra­struc­ture op­er­at­ors (four do­mains, 23 sub­do­mains, 47 main con­trols), reach­ing even to as­sets held out­side the king­dom. Ara­mco’s SACS-002 stand­ard pushes the re­quire­ment down the sup­ply chain. No vendor may re­gister without a cy­ber­se­cur­ity com­pli­ance cer­ti­fic­ate, re­newed every two years, which man­u­fac­tures a con­tinu­ous audit-and-re­medi­ation mar­ket be­neath the op­er­ator layer. Qatar’s NCSA OT re­com­mend­a­tions and the UAE’s In­form­a­tion As­sur­ance Stand­ard apply par­al­lel pres­sure.

Operational TechnologyState of Play46 •

The mech­an­ism is con­sist­ent across the re­gion, and it ex­plains why the mar­ket grows faster than the threat alone would jus­tify:

  • Man­date a con­trol, at­tach a pen­alty, and a com­pli­ance mar­ket fol­lows.
  • Re­quire third-party cer­ti­fic­a­tion (SACS-002), and a re­cur­ring ser­vices mar­ket fol­lows be­neath it.
  • Re­quire in-coun­try de­liv­ery, and a do­mestic-pro­vider mar­ket fol­lows on top.

Cap­ital builds the as­sets; the rule­book makes se­cur­ing them com­puls­ory; loc­al­isa­tion re­quires it be done loc­ally. The three en­gines run on the same fuel, which is why the de­mand side of this mar­ket is among the most pre­dict­able in tech­no­logy.

Saudi OT controls required, by facility criticality level
controls and subcontrols
05010015020056Level 3, low117Level 2, moderate151Level 1, high
Source: Saudi National Cybersecurity Authority, 2022

The sup­ply-chain en­gine is harder to size than it looks. The count of sup­pli­ers hold­ing Ara­mco’s cy­ber­se­cur­ity com­pli­ance cer­ti­fic­ate is not pub­lished, and no com­pany total was found. The only sup­plier num­ber loc­ated is a con­sultancy’s own cli­ent count. What is veri­fi­able is the ar­chi­tec­ture. It is a two-year cer­ti­fic­ate re­newed on con­tinu­ing com­pli­ance, tiered so that gen­eral vendors self-as­sess under re­mote veri­fic­a­tion while sup­pli­ers with net­work con­nectiv­ity or crit­ical-data pro­cessing face man­dat­ory on-site as­sess­ment. Three mu­tu­ally in­con­sist­ent con­trol counts cir­cu­late for the stand­ard it­self. None of them should be used.

That reg­u­la­tion lifts pre­pared­ness is, un­usu­ally, evid­enced rather than as­ser­ted. Of op­er­a­tional-tech­no­logy prac­ti­tion­ers sur­veyed in 2025, 57 per cent had a ded­ic­ated in­cid­ent-re­sponse plan for their con­trol en­vir­on­ment, rising to 70 per cent where the or­gan­isa­tion was both reg­u­lated and equipped with threat in­tel­li­gence. On the pro­vider side the Saudi li­cens­ing re­gister gives the mar­ket's clearest sig­nal of con­sol­id­a­tion. There were 16 li­censed man­aged se­cur­ity op­er­a­tions centre com­pan­ies in 2025, rising to 26 on the re­gister as at Au­gust 2026, of which only six hold the tier that per­mits ser­vice to crit­ical na­tional in­fra­struc­ture.

The structure of the Saudi controls is as stated and the scope is not.
Operational TechnologyState of Play47 •

The binding constraint is people

Capital is abundant and regulation is firm; what the Gulf lacks is the workforce to deliver. The constraint is not money but skills.

OT se­cur­ity sits at a nar­row in­ter­sec­tion (in­dus­trial-con­trol en­gin­eer­ing and cy­ber­se­cur­ity), and the tal­ent in it is scarce every­where. Glob­ally, only about 14% of or­gan­isa­tions are con­fid­ent they have the people and skills they need (WEF, 2025), and sur­veys put a lack of in­ternal re­sources as the single most-cited blocker to OT-se­cur­ity pro­grammes (SANS, 2025). Saudi Ar­a­bia coun­ted on the order of 20,000 cy­ber­se­cur­ity pro­fes­sion­als in 2024 (NCA), about 32% of them women: a strong base by global stand­ards, but far below de­mand once OT is in­cluded.

The Gulf adds its own mul­ti­pli­ers. Na­tion­al­isa­tion quotas (Saud­iz­a­tion, Emir­at­isa­tion) re­quire local hir­ing at the very mo­ment the local OT-se­cur­ity pipeline barely ex­ists, and in top-tier firms fewer than one in ten cyber roles is held by a na­tional. The skill is rarer still than IT se­cur­ity: cross-train­ing an ana­lyst into OT takes twelve to eight­een months, OT-spe­cial­ist roles can stay open for over half a year, and most OT en­gin­eers reach the field by ac­ci­dent rather than design. So the mar­ket’s growth is gated less by ap­pet­ite than by how fast the re­gion can build the people to serve it.

Key takeaways
  • The scarce input is OT-plus-se­cur­ity en­gin­eer­ing tal­ent, not cap­ital.
  • Na­tion­al­isa­tion quotas com­pound the gap: <10% of senior cyber roles held by na­tion­als.
  • Value ac­crues to who­ever builds the work­force: the pro­viders, not the pro­curers.

Unique scarcity is not what the best re­gional com­par­ison shows, and the fig­ure quoted above has been su­per­seded. On the 2026 evid­ence, 53 per cent of or­gan­isa­tions in the Middle East and North Africa say they have the people and skills they need today. That ranks the re­gion fourth of seven and ahead of North Amer­ica, 16 to 18 points clear of Sub-Saha­ran Africa and Latin Amer­ica. The pub­lisher’s own foot­note, which must travel with every fig­ure drawn from it, re­cords that the re­gional re­spond­ent count is lower than else­where and that the find­ings carry re­duced stat­ist­ical ro­bust­ness. The de­fens­ible Gulf claim is sa­li­ence rather than scarcity. On that meas­ure, 58 per cent of re­gional or­gan­isa­tions name the skills short­age as one of the biggest obstacles to re­si­li­ence, against a global 45 per cent, a gap of 13 points. That is a state­ment about am­bi­tion, not about present staff­ing.

The global-gap genre should be dropped al­to­gether. The 3.4 mil­lion fig­ure is a 2022 es­tim­ate, su­per­seded twice by its own pub­lisher and then dis­con­tin­ued in 2025 on the stated ground that prac­ti­tion­ers pri­or­it­ise crit­ical skills over head­count. For the Gulf's own re­gion the 2024 es­tim­ate was 124,978, about 2.6 per cent of the global total and twenty-seven times smal­ler than the re­cycled line. Saudi Ar­a­bia is the only Gulf state

Operational TechnologyState of Play48 •
Organisations reporting they have the cyber skills they need, 2026
% of respondents
South Asia57Europe and Central Asia57East Asia and Pacific56Middle East and North Africa53North America52Sub-Saharan Africa37Latin America and Caribbean35
Source: World Economic Forum, 2026

pub­lish­ing a work­force census at all, and its fig­ures are 19,600 spe­cial­ists in 2023, 21,300 in 2024 and more than 22,000 in 2025, with women at 32, 32 and 33 per cent.

The op­er­a­tional-tech­no­logy cut is nar­rower still and global rather than Gulf. Of prac­ti­tion­ers sur­veyed in 2025, only 9 per cent ded­ic­ated all of their time to in­dus­trial-con­trol se­cur­ity. Op­er­a­tional func­tions con­trolled the op­er­a­tional-tech­no­logy budget in about 26 per cent of or­gan­isa­tions. Of the same prac­ti­tion­ers, 65 per cent re­garded the dis­cip­line as a primary re­spons­ib­il­ity of their role, while 27 per cent led the budget. Re­spons­ib­il­ity is widely felt; au­thor­ity is not. The fig­ures quoted here for cross-train­ing time, va­cancy dur­a­tion and na­tional par­ti­cip­a­tion in cyber roles were not veri­fi­able.

The workforce build

The region’s answer to its binding constraint is to manufacture the workforce: deliberately, and at national scale.

The Gulf is not wait­ing for the tal­ent mar­ket to clear. Saudi Ar­a­bia has stood up a layered pro­gramme: the CyberIC ini­ti­at­ive (with SITE) re­port­ing sup­port for some 10,000 Saudis, Na­tional Cy­ber­se­cur­ity Boot­camps launched in 2024, a Na­tional Cy­ber­se­cur­ity Academy, and Tuwaiq Academy’s in­tens­ive courses. The UAE, for its part, runs “Cyber Sniper” and coder pro­grammes under its 2025–2031 strategy. These are gen­eral cyber pipelines, but they cre­ate the base from which OT spe­cial­ists are drawn.

The ex­pli­citly OT-fo­cused ef­forts are fewer and newer, which is ex­actly the point. They are the academy named in the Ara­mco–Dra­gos part­ner­ship, Cy­ber­ani’s train­ing part­ner­ship with KPMG to de­velop Saudi OT spe­cial­ists, and Pet­ro­leum De­vel­op­ment Oman’s work­force pro­gramme with Oman­tel. The sig­nal for the mar­ket is that the work­force is being treated as in­fra­struc­ture, built de­lib­er­ately and fun­ded by the same sov­er­eign cap­ital that funds the plants. For pro­viders, train­ing is no longer a cost centre but a route to mar­ket. The firms that build the people will own the re­la­tion­ships, which is why the na­tional cham­pi­ons lead with academies as much as with plat­forms.

Operational TechnologyState of Play49 •

Al­most every pub­lished fig­ure for the Saudi pro­gramme is a tar­get set at launch rather than an achieved in­take, and chart­ing them as in­takes would be a cat­egory error. The sup­port of some 10,000 Saudis, the 1,500 per­son­nel in na­tional au­thor­it­ies, the 150 of­fi­cials, the 5,000 to be trained through ad­vanced ex­er­cises and the 40 star­tups were all stated as ob­ject­ives in Au­gust 2022. The sub-tar­gets do not sum. Adding 1,500 plus 150 plus 5,000 gives 6,650 against a head­line of about 10,000. The only genu­ine out­comes loc­ated are a com­pet­i­tion fun­nel of 50 qual­i­fy­ing teams re­duced to 20 grant re­cip­i­ents. The reg­u­lator's own 2025 re­port­ing gives more than 13,000 be­ne­fi­ciar­ies of the pro­gramme's second phase and more than 10,000 be­ne­fi­ciar­ies of spe­cial­ist train­ing, which are the fig­ures to use.

Cybersecurity specialists in Saudi Arabia
persons
06,25012,50018,75025,00019,600202321,300202422,0002025
Source: Saudi National Cybersecurity Authority, 2026

Two of the names used above are wrong. Qatar’s gradu­ate pipeline runs through the Fresh Gradu­ates Train­ing Camp, whose twelfth co­hort launched in Au­gust 2026 along­side a first co­hort of a man­agers pro­gramme. No pro­gramme match­ing the name usu­ally given could be found, and the academy pub­lishes no par­ti­cipant counts. The Emir­ati tar­get of more than 40,000 cy­ber­se­cur­ity pro­fes­sion­als comes from the 2019 na­tional strategy, which states on its own terms that it was in force for three years. It is a lapsed tar­get rather than a plan. The one con­crete in­take fig­ure veri­fied any­where in the Gulf is a fed­er­a­tion boot­camp co­hort of 100 gradu­ates over 13 weeks, with more than 580 train­ing hours and five in­ter­na­tional cer­ti­fic­ates.

The stra­tegic read­ing sur­vives in­tact. The sup­ply side sup­ports it bet­ter than the pro­grammes do. Saudi out­put of cy­ber­se­cur­ity gradu­ates rose from about 2,500 in 2023 to more than 5,000 in 2024, while the stand­ing work­force grew by roughly 1,700 in the same year. An­nual gradu­ate out­put equal to about a quarter of the work­force, against net growth of under a tenth, im­plies sub­stan­tial leak­age or non-cyber place­ment. That is a ques­tion the data raises rather than a con­clu­sion it sup­ports, and it is the right ques­tion for any­one fund­ing an academy.

Operational TechnologyState of Play50 •
The only genuine outcomes located are a competition funnel of 50 qualifying teams reduced to 20 grant recipients.

The frontier: AI on both sides

The leading edge of Gulf OT is autonomous and AI-defended, which concentrates the market’s central tension into a single bet.

The most ad­vanced Gulf op­er­at­ors are mov­ing past mon­it­or­ing to­ward autonomy. Ara­mco and Yokogawa com­mis­sioned autonom­ous con­trol agents at a major gas fa­cil­ity in 2025. ADNOC’s AIQ runs more than 200 AI tools across its op­er­a­tions. NEOM’s Tonomus is build­ing an AI cog­nit­ive-se­cur­ity plat­form span­ning OT, with sov­er­eign-cloud of­fer­ings ex­tend­ing to in­dus­trial work­loads. The same ar­ti­fi­cial in­tel­li­gence that op­tim­ises the plant is being asked to de­fend it.

This fron­tier car­ries the mar­ket’s de­fin­ing ten­sion in con­cen­trated form. Autonomy takes hu­mans out of the con­trol loop, im­prov­ing ef­fi­ciency and re­mov­ing a layer of judge­ment in the same mo­tion. AI-driven de­fence prom­ises to close the OT-se­cur­ity and skills gaps at once (an auto­mated ana­lyst where human ones can­not be hired), but only if it ma­tures faster

than AI-en­abled at­tack widens the threat. The Gulf, build­ing green­field and fun­ded to ex­per­i­ment, is pla­cing this bet at the largest scale in the world. That makes the re­gion not only a buyer of OT se­cur­ity but a prov­ing ground for what it be­comes over the next dec­ade.

The same AI that runs the plant is being asked to defend it. The Gulf is where that bet is largest.

The fron­tier is being built as a power pro­gramme be­fore it is a com­pute pro­gramme. Saudi Ar­a­bia’s largest an­nounced cam­pus entered the pub­lic re­cord on 4 May 2026 not as a chip order but as a tender for a 380/132/33 kilo­volt net­work, 500 and 200 mega­volt-ampere sub­sta­tions and a 2,000 mega­volt-ampere bulk sup­ply point. The sov­er­eign AI vehicle it­self changed hands as this read was pre­pared. The trans­ac­tion closed on 1 Novem­ber 2026 with the sov­er­eign fund at 51 per cent and the na­tional oil com­pany at 49 per cent, for con­sid­er­a­tion of about SAR 9.9 bil­lion. The share­hold­ers’ joint state­ment gives a planned de­ploy­ment of 3 gigawatts by 2030, half the fig­ure usu­ally quoted from trade brief­ings. Every gigawatt of that is sub­sta­tions, switchgear and cool­ing plant. That is op­er­a­tional tech­no­logy.

Meas­ured ca­pa­city is a small frac­tion of the am­bi­tion and the gap is a de­liv­ery prob­lem rather than a fore­cast­ing dis­pute. Op­er­a­tional Saudi data centre ca­pa­city was 467 mega­watts in the first quarter of 2026, against 68

Operational TechnologyState of Play51 •
Operational data centre capacity, Saudi Arabia
MW, basis undefined by publisher
01252503755006820214402025467Q1 2026
Source: Saudi Ministry of Communications and Information Technology, 2026

mega­watts in 2021, a run rate of roughly 94 mega­watts a year. The na­tional 2030 tar­get re­quires about 218 mega­watts a year sus­tained for five years, and the flag­ship com­pany's own 2034 pipeline would re­quire about 700 mega­watts a year. Live AI train­ing ca­pa­city in the King­dom is ef­fect­ively nil. The op­er­at­ing base is coloca­tion, two cloud re­gions, two re­search clusters and one in­fer­ence re­gion.

On the de­fens­ive side the re­gional ap­pet­ite is high and the con­straint is the fa­mil­iar one. Of re­gional or­gan­isa­tions sur­veyed in 2026, 78 per cent had im­ple­men­ted AI-en­abled tools for cy­ber­se­cur­ity. The lead­ing hurdle to ad­opt­ing them was in­suf­fi­cient skills at 58 per cent, and the second was the need for human val­id­a­tion of AI-gen­er­ated re­sponses at 47 per cent. An auto­mated ana­lyst still re­quires an ana­lyst to sign the work.

What could go wrong

Three risks could bend the trajectory: one catastrophic, one quietly corrosive, one structural.

The mar­ket’s dir­ec­tion is un­usu­ally cer­tain, but its value can still be real­ised well or badly. Three risks set the range:

  • A safety-sys­tem at­tack that com­pletes. A Tri­ton that suc­ceeds at a Gulf en­ergy or water asset would pro­duce a phys­ical event, not a data breach. The mar­ket re­ac­tion would be a step-change in man­dated spend, on a timeline no one would choose: the re­gion’s de­fin­ing tail risk.
  • Spend that out­runs cap­ab­il­ity. A mar­ket grow­ing at 17% a year, pulled by reg­u­la­tion and cap­ital, can buy tools faster than it can de­ploy and op­er­ate them, leav­ing ex­pens­ive, half-con­figured plat­forms and a false sense of cover. The skills con­straint makes this the most prob­able dis­ap­point­ment.
  • Frag­ment­a­tion. Each Gulf state set­ting its own OT con­trols, with no uni­fied frame­work, raises the cost of multi-mar­ket op­er­a­tion and slows cross-bor­der de­liv­ery: a drag rather than a de­rail­ment.

None of these re­verses the mar­ket; each shapes how its value is cap­tured. The com­mon thread is cap­ab­il­ity: the win­ners will be those who close the cap­ab­il­ity gap rather than the pro­cure­ment gap, and the two are routinely con­fused in budget pa­pers.

The first risk ma­ter­i­al­ised dur­ing the period this read cov­ers, and it was kin­etic rather than di­gital. From early March 2026 Ir­a­nian drone strikes

Operational TechnologyState of Play52 •

dam­aged three cloud fa­cil­it­ies in the Gulf, two in the United Arab Emir­ates and one in Bahrain. On 15 Septem­ber 2026 the op­er­ator de­clared it­self un­able to re­store ac­cess to re­sources and data hos­ted ex­clus­ively in its Bahrain re­gion, the dam­age hav­ing spanned mul­tiple avail­ab­il­ity zones. Iran pub­lished a list of 29 tech­no­logy tar­gets across Bahrain, the United Arab Emir­ates, Qatar and Is­rael. The Emir­ati flag­ship AI cam­pus mas­ter plan is re­por­ted to be under re­vi­sion as a res­ult, with air de­fences and partly un­der­ground con­struc­tion being weighed, the stated cause being phys­ical se­cur­ity rather than power. The Ruwais in­dus­trial com­plex has sep­ar­ately been struck by drones. These are phys­ical events and must be la­belled as such, but they bear on the same as­sets.

The second risk is quieter. It is a mat­ter of dis­clos­ure. Not one of the six states pub­lishes a count of des­ig­nated crit­ical-in­fra­struc­ture op­er­at­ors, two pub­lish a sec­tor list, and none pub­lishes a count of en­tit­ies as­sessed or found com­pli­ant in the or­din­ary course. The re­gion's largest op­er­ator dis­closes cyber risk qual­it­at­ively and states non-ma­ter­i­al­ity while pub­lish­ing no budget line. The Gulf does not ap­pear among the re­gions rep­res­en­ted in the lead­ing an­nual global sur­vey of in­dus­trial-con­trol se­cur­ity. A mar­ket this large can­not be audited from the out­side, which is a risk an in­vestor should price rather than a gap to apo­lo­gise for.

The third is cap­ab­il­ity, and it is meas­ur­able. Of prac­ti­tion­ers sur­veyed in 2025, 31 per cent had no formal cent­ral­ised in­vent­ory, or no in­vent­ory at all, of act­ive re­mote ac­cess points into their con­trol en­vir­on­ments. An­other

60 per cent named lack of in­ternal re­sources as the blocker to com­plet­ing se­cure re­mote ac­cess. Timelines also slip as a mat­ter of course. The re­gional in­ter­con­nec­tion to Iraq was ex­pec­ted op­er­a­tional in the fourth quarter of 2025 and re­mains in pro­gress, and Kuwait's hy­per­scale sub­sta­tion con­tract was un­awar­ded at the end of 2026.

The next two to three years

What an operator or investor should expect the Gulf OT market to look like by 2028.

Over the next two to three years, the Gulf OT-se­cur­ity mar­ket will move from aware­ness to op­er­a­tion, and the centre of grav­ity will shift with it. Four de­vel­op­ments are the most likely, and the most ac­tion­able:

  • Spend mi­grates from de­tec­tion to pro­tec­tion. Hav­ing learned what they can­not see, op­er­at­ors will in­vest in se­cure re­mote ac­cess, seg­ment­a­tion and act­ive de­fence, rather than fur­ther vis­ib­il­ity tools.
  • The na­tional cham­pi­ons con­sol­id­ate the re­cur­ring layer. Sov­er­eign-backed pro­viders (Cy­ber­ani, SITE, Help AG, CPX) will cap­ture more of the man­aged-ser­vice mar­ket as loc­al­isa­tion rules bite, sit­ting between op­er­at­ors and the global spe­cial­ists.
Operational TechnologyState of Play53 •
  • Reg­u­la­tion tight­ens and spreads. Saudi Ar­a­bia’s OT con­trols set the tem­plate; ex­pect the UAE, Qatar and even­tu­ally the smal­ler states to deepen OT-spe­cific man­dates, each adding re­cur­ring spend.
  • AI de­fence is tested in earn­est. The giga-pro­jects and the most ad­vanced op­er­at­ors will move autonom­ous, AI-driven OT de­fence from pilot to pro­duc­tion: the de­cis­ive ex­per­i­ment of the period.

The bind­ing ques­tion through­out is the work­force. The mar­ket can fund and man­date OT se­cur­ity faster than it can staff it, so the op­er­at­ors and pro­viders that solve the people prob­lem (through academies, man­aged ser­vices and AI aug­ment­a­tion) will define the win­ners by 2028.

Key takeaways: the outlook
  • Spend shifts de­tec­tion → pro­tec­tion; se­cure re­mote ac­cess and seg­ment­a­tion lead.
  • Na­tional cham­pi­ons con­sol­id­ate the re­cur­ring man­aged-ser­vice layer.
  • The de­cis­ive test is AI de­fence, pilot to pro­duc­tion: gated by the work­force.

The most use­ful dis­cip­line for a two-to-three-year view is to sep­ar­ate what is fin­anced from what is an­nounced, be­cause in this mar­ket the two dif­fer by roughly a factor of three. On the grid, three ex­pan­sions are under con­tract rather than under dis­cus­sion. The Kuwait in­ter­con­nec­tion reached com­ple­tion in July 2025 at a total pro­ject value of US$260 mil­lion. The Emir­ati ex­pan­sion from 2,400 to 3,500 mega­watts is fin­anced with US$205 mil­lion from the Abu Dhabi Fund for De­vel­op­ment. The dir­ect Omani link,

1,700 mega­watts of trans­mis­sion ca­pa­city and 1,200 mega­watts net trans­fer ca­pa­city, car­ries US$500 mil­lion of trans­itional fin­an­cing from a com­mer­cial bank plus an un­dis­closed Qatari tranche. One fur­ther cor­ridor, Ghunan to Jasra, is ap­proved for im­ple­ment­a­tion. A second Saudi con­verter sta­tion and the Qatari cor­ridor re­main stud­ies only.

GCC interconnection capacity, built up from per-link figures
MW
03,7507,50011,25015,0005,050Pre-20258,350After July 202510,250On completion
Source: GCC Interconnection Authority, 2026

On the mar­ket side, con­sol­id­a­tion should be ex­pec­ted to show up in re­gisters rather than in ana­lyst tables. Li­censed man­aged se­cur­ity op­er­a­tions centre pro­viders in Saudi Ar­a­bia rose from 16 in 2025 to 26 in Au­gust 2026, while the tier per­mit­ting ser­vice to crit­ical na­tional in­fra­struc­ture re­mains held by six firms. Re­gistered pro­viders rose from 353 at end-2023 to roughly 525 in Oc­to­ber 2026. Site-level cer­ti­fic­a­tion against the in­ter­na­tional stand­ard be­came avail­able only in Feb­ru­ary 2026, so any Gulf site

Operational TechnologyState of Play54 •

cer­ti­fic­ate will be newly is­sued. The global cer­ti­fied pop­u­la­tion re­mains small enough to count.

The bind­ing work­force ques­tion is also the one most likely to re­solve slowly and vis­ibly. Saudi out­put of cy­ber­se­cur­ity gradu­ates more than doubled between 2023 and 2024 while the stand­ing work­force grew about 5.9 per cent a year across 2023 to 2025. Those two rates are not re­con­cil­able with a mar­ket that in­tends to triple its man­aged-ser­vice base, and the gap will be closed either by academies, by man­aged ser­vices, by auto­ma­tion or not at all.

The read, on a page

The argument, in one place, with the evidence behind it.

The Gulf OT-se­cur­ity mar­ket is small, fast-grow­ing, and un­usu­ally cer­tain in its dir­ec­tion, be­cause the forces driv­ing it are struc­tural rather than cyc­lical. The air gap that pro­tec­ted the re­gion’s in­dus­trial sys­tems is gone, and its dis­ap­pear­ance cre­ated both the mar­ket and the risk in the same mo­tion. The de­mand is built in: an enorm­ous, di­git­ising in­dus­trial base; a threat his­tory no mar­ket­ing could in­vent; sov­er­eign cap­ital that builds the as­sets; reg­u­la­tion that man­dates their de­fence; and an en­ergy trans­ition that mul­ti­plies the sur­face.

The mar­ket will com­pound through the dec­ade, led by Saudi Ar­a­bia and the en­ergy sec­tor, with the giga-pro­jects as its green­field show­case and AI-de­fen­ded autonomy as its fron­tier. The ques­tion is not whether it grows but who cap­tures the value, and the an­swer turns on cap­ab­il­ity, not cap­ital. The win­ners will be the pro­viders that can de­liver, in-coun­try, the scarce en­gin­eer­ing the work re­quires, and the op­er­at­ors that treat OT se­cur­ity as an op­er­at­ing dis­cip­line rather than a pro­cure­ment line.

The market’s growth is certain; its value accrues to whoever solves the people problem, not whoever spends the most.

For a re­gion re­build­ing its eco­nomy on con­nec­ted in­dustry, se­cur­ing the op­er­a­tional layer is not a cost of di­git­isa­tion. It is the pre­con­di­tion for it.

Three cor­rec­tions should travel with this sum­mary, be­cause the stand­ard ver­sion of the ar­gu­ment is wrong by an order of mag­nitude in each case. The Gulf’s en­tire in­ter­net-ex­posed in­dus­trial foot­print is 100 unique IPv4 ad­dresses. The fig­ure is not in the thou­sands. The case for the re­gion there­fore rests on the com­pos­i­tion of its ex­pos­ure and on the bound­ary a pub­lic scan can­not see, not on volume. The US$500 bil­lion at­tached to one giga-pro­ject is a 2017 an­nounce­ment, against an audited car­ry­ing value of about US$56.2 bil­lion for all five giga-pro­ject com­pan­ies com­bined at end-2024. And India, the com­par­ator in­dus­trial eco­nomy most

Operational TechnologyState of Play55 •

often held up as a warn­ing, runs seven times the Gulf’s ex­posed ad­dress count with no con­firmed com­prom­ise of a con­trol sys­tem any­where in its pub­lic re­cord.

Saudi cybersecurity provider counts, by register
companies
Registered providers, Oct 2026525Registered providers, end-2023353Licensed MSOC, Aug 202626Licensed MSOC, 202516Tier 1 licences, Aug 20266
Source: Saudi National Cybersecurity Authority, 2026

The meas­ure­ment prob­lem is not in­cid­ental to the ar­gu­ment. It is part of it. The most rig­or­ous na­tional meas­ure­ment of cyber spend­ing in the re­gion, pub­lished an­nu­ally by a reg­u­lator with a stated tax­onomy, con­tains no op­er­a­tional-tech­no­logy break­out in any of its three edi­tions. No named ana­lyst house pub­lishes a vendor-level share for this mar­ket, so it can be ranked po­s­i­tion­ally and sized in ag­greg­ate but not di­vided up. No Gulf state pub­lishes a count of the crit­ical-in­fra­struc­ture op­er­at­ors it reg­u­lates. The com­pet­it­ive in­tel­li­gence in this read is there­fore drawn from re­gisters, li­cences, con­tracts and audited seg­ment notes, which is a nar­rower evid­ence base than ana­lyst tables but a sounder one.

None of this changes the con­clu­sion. De­mand is struc­tural, the rule­book is bind­ing, the as­sets are being built, and the scarce input is the en­gin­eer­ing ca­pa­city to de­liver in-coun­try. What an op­er­ator ac­tu­ally con­trols is the se­quence: in­vent­ory first, then seg­ment­a­tion, then mon­itored re­mote ac­cess.

Key takeaways

Nine conclusions for the people who must act on this market.

  1. The air gap is gone. The Gulf’s in­dus­trial OT is con­nec­ted; isol­a­tion is no longer the se­cur­ity model, and the pos­ture must change with it.
  2. The op­por­tun­ity and the risk are one con­nec­tion. Mod­ern­isa­tion and ex­pos­ure are the same links viewed twice: you can­not pur­sue one without man­aging the other.
  3. The mar­ket is struc­tur­ally cer­tain. Four en­gines (di­git­isa­tion, threat lin­eage, sov­er­eign cap­ital, reg­u­la­tion) all push the same way.
  4. OT is not IT. Un­patch­able, al­ways-on, phys­ical-con­sequence sys­tems re­quire a spe­cial­ist dis­cip­line, not the IT toolkit ap­plied to a fact­ory.
  5. The threat is local and proven. Shamoon and Tri­ton happened here; the re­gion buys on memory, and the cam­paign is now con­tinu­ous.
  6. Ar­chi­tec­ture beats patch­ing. Seg­ment­a­tion under IEC 62443 is the core con­trol, and ret­ro­fit­ting it is where the multi-year spend goes.
Operational TechnologyState of Play56 •
  1. Reg­u­la­tion man­u­fac­tures re­cur­ring spend. Saudi OT con­trols and Ara­mco’s sup­ply-chain stand­ard con­vert good prac­tice into com­puls­ory rev­enue.
  2. The na­tional cham­pi­ons own the fu­ture. Loc­al­isa­tion rules fa­vour sov­er­eign-backed pro­viders who hold the re­cur­ring man­aged-ser­vice layer.
  3. People are the bind­ing con­straint. The mar­ket can fund and man­date OT se­cur­ity faster than it can staff it; cap­ab­il­ity de­cides the win­ners.

Nine con­clu­sions sur­vive the evid­ence, and they are not the nine most Gulf com­ment­ary of­fers. First, ex­pos­ure is a ques­tion of com­pos­i­tion rather than volume. The six states hold about one eighth of one per cent of the world's in­ter­net-ex­posed in­dus­trial ad­dresses, while hold­ing 1.75 per cent of ex­posed DNP3 and 1.17 per cent of the inter-con­trol-centre pro­tocol. Second, the mar­ket is smal­ler and worse meas­ured than its head­line fig­ures imply, and the ser­vices-in­clus­ive and plat­form-only siz­ings dif­fer by about five­fold. Third, Saudi Ar­a­bia has man­dated its own op­er­a­tional-tech­no­logy con­trols, not the in­ter­na­tional stand­ard, and the two are re­lated by an annex rather than by equi­val­ence. Fourth, the con­trols scale with fa­cil­ity crit­ic­al­ity on a lad­der of 151, 117 and 56, and crit­ic­al­ity is as­signed by the op­er­ator, which is why no count of reg­u­lated en­tit­ies ex­ists any­where in the re­gion.

Fifth, the li­cence rather than the brand is the bar­rier to entry: six firms hold the Saudi tier that per­mits ser­vice to crit­ical na­tional in­fra­struc­ture, against twenty in the tier that ex­cludes it. Sixth, the spe­cial­ists are loc­al­ising rather

than ex­port­ing, every re­cent Gulf agree­ment pair­ing a for­eign cap­ab­il­ity with a local fa­cil­ity, a train­ing com­mit­ment and in­creas­ingly loc­ally hos­ted in­fra­struc­ture. Sev­enth, the cap­ital that drives de­mand turned down in 2025, Gulf con­tract awards fall­ing about 32 per cent and the Emir­ates over­tak­ing Saudi Ar­a­bia as the largest pro­jects mar­ket in the first quarter of 2026.

Eighth, the work­force con­straint is one of sa­li­ence rather than scarcity. The re­gion re­ports mid-field skills ad­equacy while nam­ing the skills short­age as an obstacle to re­si­li­ence 13 points more often than the global av­er­age. Ninth, and most con­sequen­tial for any­one act­ing on this read, the con­trols that fail are ar­chi­tec­tural and ac­cess-re­lated rather than en­d­point-re­lated. The se­quence that works starts with in­vent­ory and seg­ment­a­tion, then mon­itored re­mote ac­cess, then everything else.

Volge's expertise in operational technology

The engineering work the sector asks for, and the form it takes on an engagement.

In­dus­trial con­trol sys­tems are now con­nec­ted products under reg­u­lat­ory su­per­vi­sion, so the en­gin­eer­ing they need sits across se­cur­ity, op­er­a­tions and safety at once. The work below is what op­er­at­ors, in­dus­trial

Operational TechnologyState of Play57 •

groups and their vendors in this mar­ket com­mis­sion. Each is de­livered by senior en­gin­eers work­ing in­side the cli­ent's own team, with the Global De­liv­ery Centres run­ning what has been built once it is in pro­duc­tion.

Securing the plant
  • Op­er­a­tional-tech­no­logy se­cur­ity as­sess­ment against IEC 62443, cov­er­ing zones, con­duits and the con­trols at each level.
  • Asset dis­cov­ery and in­vent­ory across the plant floor, in­clud­ing devices no cor­por­ate tool can see.
  • Seg­ment­a­tion design and build between the cor­por­ate net­work and the pro­cess net­work.
  • Se­cure re­mote ac­cess for vendors and in­teg­rat­ors, re­pla­cing shared cre­den­tials and open jump hosts.
Running it
  • Op­er­a­tional-tech­no­logy se­cur­ity op­er­a­tions centre design, build and run, with pro­tocol-aware mon­it­or­ing.
  • Vul­ner­ab­il­ity and patch man­age­ment se­quenced around main­ten­ance win­dows rather than against them.
  • In­cid­ent re­sponse play­books writ­ten for plant con­di­tions, and the ex­er­cises that test them.
  • Safety in­stru­mented sys­tem re­view, where the se­cur­ity case and the safety case meet.
Making the data useful
  • His­tor­ian and time-series in­teg­ra­tion, bring­ing pro­cess data into the en­ter­prise without open­ing a path back.
  • Con­di­tion mon­it­or­ing and pre­dict­ive main­ten­ance built on that data.
  • Di­gital twin and sim­u­la­tion for pro­cess and con­trol change.
  • Re­si­li­ence test­ing, in­clud­ing re­cov­ery of a con­trol sys­tem from cold.
How an engagement runs
  • A dia­gnosis of two to eight weeks, led on site by senior ar­chi­tects, es­tab­lish­ing where the plant and its con­trols ac­tu­ally stand.
  • A se­quenced pro­gramme, be­cause in­vent­ory has to pre­cede seg­ment­a­tion and seg­ment­a­tion has to pre­cede mon­itored re­mote ac­cess.
  • Build teams em­bed­ded along­side the cli­ent's own en­gin­eers rather than work­ing to a spe­cific­a­tion at arm's length.
  • Run and op­er­ate from the Global De­liv­ery Centres, with the run­books and the evid­ence an aud­itor will ask for.
Operational TechnologyState of Play58 •
The Read

The market’s growth is certain; its value accrues to whoever solves the people problem, not whoever spends the most.

Sources

The bodies cited in this read, with the year the text attributes to each.

Operational TechnologyState of Play59 •
Version · VSP-OTG-005 · 2026
volgeEnable Powerful
Transformations
About Volge

Volge is an engineering firm providing enterprise technology transformation services to clients across the Middle East, India, APAC, Europe, and North America.

Delivery stands on two pillars: Forward Deployed Engineering, with senior specialists embedded in the client’s own team, and Intelligent Command Centre Operations, run from our Global Delivery Centres, which keeps what they build running. Senior engineers stay with an engagement from the first diagnosis through to production.

Our services span Strategy, Data, Platforms, AI and Automation, Cyber Trust, Technology Operations, and Applied Functional Engineering. Engagements begin with a diagnosis. The firm is built on working autonomy: engineers decide, innovate, co-create, and own their work. Motivated engineers build the systems that sustain our clients’ growth.

Terms of use

This edition may be shared with organisations evaluating a Volge engagement. It may not be altered or republished without Volge’s permission, and figures carry their stated basis and date. Volge, Intelligent Command Centre, and Intelligent Command Centre Operations are marks of Volge; other names are the trademarks of their respective owners.