Malaysia AI Market 2026: Market Size, Investment and Adoption

Malaysia’s artificial intelligence market reached US$1.13 billion in 2025 and is forecast to hit US$17.4 billion by 2034, growing at a 34.4% CAGR from 2026 (IMARC Group).[1] That headline number understates what is actually happening. Between H2 2024 and H1 2025, Malaysia pulled in US$759 million in AI funding — 32% of all AI investment in Southeast Asia, according to the e-Conomy SEA 2025 report by Google, Temasek and Bain & Company.[2]

The market now has all the ingredients: a national AI agency, AI Malaysia Bhd, coordinating policy, a RM2 billion sovereign AI cloud funded in Budget 2026, a homegrown frontier-chasing model in YTL AI Labs’ ILMU, and enterprise heavyweights — Maybank, Petronas, CIMB — committing serious capital.[3][4] Malaysia’s Ministry of Digital told Parliament in December 2025 that AI could add RM13–20 billion to GDP annually by 2030 and create 300,000–500,000 high-skilled jobs.[5]

This piece covers the market: sizing, adoption, domestic players and policy. For the infrastructure story — gigawatts, GPUs and who is building what — see our Malaysia compute hub deep-dive.

How big is Malaysia’s AI market in 2026?

Depends which layer you measure. The core AI software and services market was worth US$1.13 billion in 2025, per IMARC Group, with the firm projecting 34.43% annual growth through 2034.[1] Statista’s Market Insights tracks a similar trajectory for Malaysia’s AI market through 2030, though its country-level figures sit behind a paywall.

The investment picture is more striking than the revenue one. Malaysia’s US$759 million in AI funding over the twelve months to mid-2025 put it among Southeast Asia’s top AI investment destinations (e-Conomy SEA 2025).[2] But the money is concentrated: H1 2025 saw just 23 deals, far below the 236 recorded at the 2021 peak, meaning a handful of large infrastructure and enterprise rounds are doing the heavy lifting.[2] Consumer monetisation is moving too — revenue from apps marketing AI features in Malaysia grew 103% year on year in H1 2025, per the same report.[2]

The 2026 edition of the same report shows adoption catching up with the money. Malaysia ranks second in Southeast Asia for AI adoption, with 68% of the working population using generative AI, and enterprise workloads account for 45–60% of AI usage in Malaysia and Singapore (e-Conomy SEA 2026).[6]

On the macro side, the Ministry of Digital’s December 2025 parliamentary estimate — 0.8–1.2% of additional GDP from AI by 2030, worth RM13–20 billion a year — is the government’s own baseline for what the market is playing for.[5] For how Malaysia’s numbers stack against China, Japan, India and the rest of the region, see our Asia AI market country breakdown.

What is Malaysia’s AI policy in 2026?

Malaysia now has one of the most developed AI policy stacks in Southeast Asia. The National AI Office (NAIO), set up under the Ministry of Digital, has handed delivery of the “AI Nation by 2030” agenda to AI Malaysia Bhd, which is running the National AI Action Plan 2026–2030 and its 28 initiatives and leading work on a Malaysia AI Safety Institute.[7][8] The missing piece is law. NAIO consulted on an AI Governance Bill in July 2026, proposing a Central AI Authority and a three-tier, risk-based regime that could also reach foreign providers whose AI systems are used in Malaysia (Baker McKenzie, July 2026).[9] Digital Minister Gobind Singh Deo has said the draft is going to Cabinet, with tabling in Parliament expected by early 2027 (The Star, September 2026).[10] Until then, Malaysia has no law that specifically governs AI.

The RM2 billion sovereign AI cloud is real — and MCMC is building it

The single biggest policy commitment is the sovereign AI cloud: RM2 billion (roughly US$475 million), announced in the Budget 2026 speech on 10 October 2025, to be developed by the Malaysian Communications and Multimedia Commission (MCMC).[3] The stated goal is to keep Malaysian government and enterprise data — and the value generated from it — inside national borders (Sinar Daily, October 2025).[3] Prime Minister Anwar Ibrahim restated the sovereignty line in April 2026, saying “the rules will be made in Malaysia” and that Malaysia does not intend to remain merely a consumer of AI (New Straits Times, April 2026).[11] Budget 2026 also included RM5.9 billion across ministries for R&D and commercialisation, plus an additional 50% tax deduction for small and medium businesses’ spending on accredited AI and cybersecurity training.[12]

Budget 2027, tabled on 9 October 2026, built on that rather than adding a new headline project: RM6 billion for R&D and commercialisation across ministries, close to RM15 million for AI Malaysia Bhd tied to a target of 200,000 AI-skilled workers, RM30 million for MDEC to help 4,000 small businesses adopt AI and certify 5,000 AI professionals, and another 100,000 free AI subscriptions, ChatGPT included, for young Malaysians (The Star; Malay Mail; ProductNation, October 2026).[13][14][15]

Notably, Malaysia has not restricted access to foreign AI models — ChatGPT, Claude and Gemini all operate freely, as we cover in our guide to AI access in Malaysia. The sovereignty play is about infrastructure and data residency, not model bans.

ILMU: Malaysia actually built its own AI model

Most national LLM projects in Southeast Asia are fine-tunes. ILMU, from YTL AI Labs (a unit of YTL Power International), is the exception with genuine momentum. Unveiled in August 2025 as a fully multimodal model handling text, voice and vision, ILMU topped the Malay MMLU benchmark — beating GPT-4o, DeepSeek V3 and GPT-5 on Bahasa Melayu tasks, according to YTL AI Labs’ launch benchmarks (W.Media, September 2025).[16] ILMUchat, the consumer app, opened early access on 16 September 2025 — Malaysia Day.[4]

Progress has continued into 2026. On 21 April 2026, YTL AI Labs launched ILMU Claw, an agent-building platform on the OpenClaw framework, powered by ILMU-Nemo-Nano — a model co-developed with NVIDIA — and hosted entirely on Malaysian infrastructure via YTL AI Cloud, with 800 seats at launch (YTL Power International, April 2026).[17] “We are excited to continue giving Malaysian builders what they deserve — world-class AI capability, on home soil,” said YTL AI Labs CEO Foong Chee Mun. For startups, an ILMU Accelerator Programme run with MDEC offers RM5 million in free ILMU API credits.[4] A sovereign model, running on sovereign compute, powering agents: that is a stack no other Southeast Asian country outside Singapore can currently claim.

How widely is AI adopted in Malaysia?

Consumer adoption is among the highest in Asia. 67% of Malaysian adults used an AI tool within the previous three months, per the Vodus AI Consumption Study (n=2,556, fieldwork December 2025–January 2026), and 74% of digital consumers interact with AI daily (e-Conomy SEA 2025).[18][2] Sentiment matches usage: Ipsos ranked Malaysia third globally for AI enthusiasm at 77% in 2025, with 73% trusting the government to regulate AI — against just 31% in the US.[19][20]

Enterprise is the lagging layer — with a familiar gap between use and value. 38% of Malaysian businesses were consistently using AI by mid-2026, up from 27% a year earlier, with financial services (53%) and manufacturing (50%) leading, per AWS’s “Unlocking Malaysia’s AI Potential 2026” study.[21] But 67% of adopters still use it only for basic tasks, and just 19% have a formal strategy to scale it across the business (AWS, 2026).[21] Meanwhile BCG’s “AI at Work” research found 79% of Malaysian workers already use AI tools regularly, above the 72% global average.[22] Employees are ahead of their employers, and the 2026 enterprise opportunity is closing that gap.

Which companies are driving Malaysia’s AI market?

Three clusters matter. First, the conglomerates: YTL is the standout, pairing YTL AI Labs’ models with its US$4.3 billion NVIDIA-partnered AI data centre in Johor.[23] Petronas is the deepest enterprise deployer — at ADIPEC in November 2025, its upstream arm Malaysia Petroleum Management announced AI and high-performance computing collaborations with SLB, Shell, TotalEnergies, ConocoPhillips and Microsoft, spanning seismic imaging, agentic AI and remote autonomous offshore operations (Petronas, November 2025).[24]

Second, the banks. Maybank’s ROAR30 strategy, announced 20 January 2026, commits RM10 billion to technology, data and AI over 2026–2030, covering AI-driven credit assessment, fraud detection and a next-generation app (Fintech News Malaysia, January 2026).[25] CIMB and RHB have piloted AI agent-initiated payments with Mastercard, completing Malaysia’s first live agentic transactions in March 2026 (Lowyat.NET, March 2026).[26]

Third, the startup layer — the weak spot. The funding concentration noted above (23 deals in H1 2025) reflects a thin domestic venture pipeline; the government’s answer is the NAIO, MDEC support, Budget 2026’s tax incentives and the ILMU Accelerator.[2] The Johor–Singapore Special Economic Zone (JS-SEZ) adds a cross-border dimension: at the second joint investment forum on 14 October 2025, Singapore-based firms committed more than S$5.5 billion into Johor (Singapore MTI; Malay Mail, October 2025).[27][28] Malaysia is targeting 50 projects in the zone’s first five years, 100 within ten, and 20,000 skilled jobs within five years (The Star, July 2026).[29] Expect AI services firms to use the SEZ to arbitrage Singapore demand against Malaysian cost.

Compute: the constraint under the market

The market runs on Malaysia’s data centre boom — covered in depth in our Malaysia compute hub deep-dive — so here is the short version. Johor alone has grown from 10MW of data centre capacity at the start of 2021 to around 1.3GW by November 2024 (ISEAS).[30] Malaysia led Southeast Asia with 1.04GW under construction in the first half of 2026 (Cushman & Wakefield, via Data Center Knowledge),[31] and Johor is projected to hold 60% of national capacity by 2030 (The Vibes, June 2025).[32] NVIDIA-YTL, Microsoft and ByteDance (a planned US$2.1 billion AI investment) anchor the pipeline — part of the wider regional build-out mapped in our SEA data centre boom tracker.[23][33]

But 2026 marked the end of the open-door era. On 24 February 2026, PM Anwar confirmed in Parliament that applications for data centres unrelated to AI and high technology had been halted — with curbs in place for around a year and a half — citing power and water strain (Malay Mail, February 2026).[34] The same month, Malaysia saw its first data centre protest: about 50 people in Gelang Patah, Johor, rallied against a Zdata Technologies facility being built less than a kilometre from homes, citing dust, water supply and an earlier stop-work order (W.Media, February 2026).[35] Social licence is now a real variable in Malaysia’s AI build-out.

Where does Malaysia sit in the US–China chip fight?

Squarely in the middle — and so far, playing it well. Since July 2025, Malaysia has required permits for the export or trans-shipment of US-origin high-end AI chips, a move designed to reassure Washington that Nvidia GPUs landing in Malaysian data centres stay there (Bloomberg, July 2025).[36] The pressure tightened on 31 May 2026, when the US Commerce Department clarified that AI chip licensing requirements apply to Chinese-headquartered firms anywhere in the world — closing the overseas-subsidiary loophole and putting Chinese-owned data centre tenants in Malaysia directly in scope (Reuters, June 2026).[37] That makes tenant mix a strategic question for every Malaysian operator. The full trade-and-supply-chain picture is in our analysis of US–Malaysia semiconductors and data centres.

Malaysia AI Market 2026: FAQ

How big is Malaysia’s AI market?

US$1.13 billion in 2025, forecast to reach US$17.4 billion by 2034 at a 34.43% CAGR (IMARC Group).[1] Malaysia also attracted US$759 million in AI funding in the year to mid-2025 — 32% of Southeast Asia’s total (e-Conomy SEA 2025).[2]

What is Malaysia’s National AI Office (NAIO)?

NAIO is the Ministry of Digital unit that laid Malaysia’s AI governance groundwork, including the 2026 consultation on an AI Governance Bill.[9] Delivery of the National AI Action Plan 2026–2030 is now coordinated by AI Malaysia Bhd, which is also setting up a Malaysia AI Safety Institute.[8]

What is ILMU?

ILMU is Malaysia’s homegrown multimodal AI model, built by YTL AI Labs. It topped the Malay MMLU benchmark against GPT-4o and DeepSeek V3 at its August 2025 launch, and in April 2026 expanded into AI agents with ILMU Claw, co-developed with NVIDIA.[4][17]

What share of Malaysian businesses use AI?

38% of Malaysian businesses were consistently using AI by mid-2026, up from 27% a year earlier (AWS).[21] Consumer adoption is higher: 67% of adults used an AI tool in the previous three months (Vodus, January 2026).[18]

Has Malaysia banned new data centres?

Not entirely. In February 2026 PM Anwar confirmed a freeze on applications for non-AI, non-high-tech data centres over power and water concerns.[34] AI-focused projects continue to be approved.

Can you use ChatGPT and Claude in Malaysia?

Yes — Malaysia imposes no restrictions on foreign AI models. See our full guide to AI access in Malaysia.

Share this article

Sources & Further Reading

  1. Malaysia Artificial Intelligence Market Size, Share and Forecast — IMARC Group, 2026
  2. e-Conomy SEA 2025 — Google, Temasek and Bain & Company, 2025
  3. MCMC to build RM2 billion ‘sovereign AI cloud’ under Malaysia’s AI Nation 2030 vision — Sinar Daily, 2025-10
  4. YTL AI Labs Launches ILMU — 100% Malaysian AI, Built by Malaysians for Malaysians — YTL Power International, 2025-08
  5. AI could add up to RM20b to Malaysia’s economy by 2030, says Digital Ministry — Malay Mail, 2025-12-04
  6. e-Conomy SEA 2026: Malaysia — Google, Temasek and Bain & Company, 2026
  7. National AI Action Plan 2026–2030 Fact Sheet — National AI Office (ai.gov.my), 2026
  8. Govt launches AI Malaysia Berhad — Lowyat.NET, 2026
  9. Malaysia: Public Consultation on the AI Governance Bill — Baker McKenzie, 2026-07-13
  10. AI Governance Bill expected by early 2027, says Gobind — The Star, 2026-09-11
  11. ‘The rules will be made in Malaysia’: Anwar draws the line on AI sovereignty — New Straits Times, 2026-04
  12. MDEC: Budget 2026 Sets Malaysia On Course To Become An AI Nation By 2030 — Business Today, 2025-10-17
  13. Budget 2027: RM6bil allocated to boost research, development and commercialisation across ministries — The Star, 2026-10-09
  14. Budget 2027: Digital Ministry targets 200,000 AI-skilled individuals in push for AI nation by 2030 — Malay Mail, 2026-10-11
  15. Budget 2027: free AI subscriptions, ChatGPT included — ProductNation, 2026-10
  16. ILMU: Deep dive into Malaysia’s first homegrown LLM with Prof Chan — W.Media, 2025-09-25
  17. YTL AI Labs Launches ILMU Claw – Malaysia’s AI Powers AI Agents — YTL Power International, 2026-04-21
  18. AI Usage in Malaysia 2026: Adoption, Trust and What’s Next — Vodus, 2026-01-20
  19. AI Monitor 2025 (press release) — Ipsos Malaysia, 2025-06
  20. Excited or Anxious? Where Malaysia Stands on AI in 2025 — Bernama, 2025-06-11
  21. 38% of Malaysian businesses use AI, but most stuck on basics — TechNode Global, 2026-08-19
  22. AI adoption in Malaysia: transforming optimism into value — The Star, 2025-12-08
  23. YTL and NVIDIA Unveil $4.3 Billion AI Infrastructure Boost in Malaysia — W.Media
  24. PETRONAS strengthens global collaboration on AI integration ecosystem and high-performance computing — PETRONAS, 2025-11
  25. Maybank ROAR30 strategy: RM10 billion technology investment — Fintech News Malaysia, 2026-01
  26. Mastercard Completes First Live Agentic AI Transaction With CIMB And RHB Pilot — Lowyat.NET, 2026-03-11
  27. 2nd Johor-Singapore Special Economic Zone (JS-SEZ) Joint Investment Forum sets the stage — Ministry of Trade and Industry Singapore, 2025-10-14
  28. Singapore firms commit over S$5.5b into Johor since JS-SEZ deal, says DPM Gan — Malay Mail, 2025-10-14
  29. Govt confident of exceeding JS-SEZ target of 20,000 jobs within five years — The Star, 2026-07-01
  30. Data Centres, Energy Demand and Sustainability: Can Malaysia Strike the Right Balance? — ISEAS – Yusof Ishak Institute, 2025-06-12
  31. APAC Data Center Pipeline Surges as AI Pre-Leasing Tightens Supply — Data Center Knowledge, 2026-08-17
  32. Johor to host 60 per cent of Malaysia’s data centre capacity by 2030: Tengku Zafrul — The Vibes
  33. ByteDance Plans $2.1B AI Investment in Malaysia, Zafrul Says — Bloomberg Law, 2024-06
  34. Malaysia freezes new non-AI data centres over power and water concerns, says Anwar — Malay Mail, 2026-02-24
  35. Johor’s data center protest, a first in Malaysia — W.Media, 2026-02-09
  36. Malaysia Controls AI Chip Exports as US Targets China Smuggling — Bloomberg, 2025-07-14
  37. US to halt Nvidia AI chip shipments to Chinese firms outside China — The Jakarta Post (Reuters), 2026-06-01
Tom Simpson

Tom Simpson

Tom Simpson is an operator, investor and writer working where AI, media, gaming and culture meet. He is the founder of AK3R, a boutique venture fund and advisory business backing and building companies in AI, agents, gaming, advertising and digital marketplaces, and writes the Hyperfuture Memo on how AI is reshaping markets, media and culture. He has spent around 20 years building media, advertising and gaming businesses across the UK, the US and Asia, including building AdColony’s Asia business through its exit, and founded Digital in Asia in 2013.

The DIA Briefing

Asia’s digital markets, every other week. Free.