Malaysia plans to introduce its first dedicated AI law by early 2027

Malaysia plans to give artificial intelligence its own rulebook. The country is preparing to introduce its first dedicated AI law in early 2027.

Digital Minister Gobind Singh Deo announced on September 10, 2026 that his ministry has finalized the draft Artificial Intelligence (AI) Governance Bill. The draft will soon go to the Cabinet. The plan is to table it in Parliament by the end of Q1 2027 at the latest.

What the AI Governance Bill would do

The bill is built on a risk-based model. AI systems would not be treated as one big category. Each system would be sorted by how much damage it could plausibly cause.

The bill sets out three tiers.

Tier 1 covers unacceptable risk. AI uses in this bucket would be prohibited where there is intent to harm.

Tier 2 covers high-risk systems. These would require assessments, oversight and documentation.

Tier 3 covers low-risk applications. These would face baseline compliance.

The draft also lays out five core principles meant to guide how AI is built and deployed:

• Human Dignity
• Transparency and Explainability
• Accountability
• Safety and Security
• Data Governance

The stated goal is to build trust in AI systems.

A new regulator enters the chat

The bill would also establish a Central AI Authority. Its planned functions span safety, investigation and enforcement, and enablement.

The enforcement structure is being set up at the same time as the bill moves forward.

The legislation is designed to complement existing laws rather than replace them. One of those is the Personal Data Protection Act from 2010, which already governs how personal information is handled.

The draft was shaped by a public consultation that ran from July 10 to July 31, 2026. Stakeholders contributed during that window, and the ministry says the result aims to improve industry accountability and transparency.

How Malaysia got here

The bill is the latest step in a steady build-up rather than a sudden pivot. Malaysia released its National Guidelines on AI Governance and Ethics in 2024.

In March 2026, the country launched the MY-AI Standards platform with SIRIM Berhad. That platform focuses on technical standards.

There is also the National AI Action Plan 2026-2030, which frames the government’s broader ambitions for the technology. The new bill sits on top of all three.

Malaysia’s governance strategy also leans on technical standards through MY-AI Standards and on faster regulatory measures that do not require a full act of Parliament.

What this means for businesses and investors

For companies building or deploying AI in Malaysia, the biggest change is clarity. Guidelines are suggestions; laws come with obligations and, eventually, consequences.

Firms working on systems likely to land in Tier 2 have the most homework. Risk assessments, documentation and human oversight all cost time and money.

Research on the bill suggests clear guidelines could boost confidence for investors and startups, since they would be operating inside a legally recognized framework.

A single Central AI Authority may also streamline compliance, potentially lowering barriers for companies looking to enter the AI space.

Malaysia’s move parallels regulatory activity in jurisdictions like the EU and the US, and the research indicates this could position the country as an attractive hub for AI development in Southeast Asia.

Several questions remain. The first is how the Cabinet handles the draft, and whether Parliament sticks to the Q1 2027 target. The second is where exactly the lines between tiers get drawn. The third is how the Central AI Authority balances its dual role of policing and promoting the industry. Finally, there is the interaction with the 2010 data protection law and how the two regimes overlap.