Malaysia is about to put its first real rulebook for artificial intelligence on paper, and one part of it could decide who legally owns the work that AI produces for you.
The country's first dedicated AI Governance Bill is due to reach the Cabinet this month, in June 2026, according to Digital Minister Gobind Singh Deo. He told the Dewan Rakyat in late 2025 that he hoped to present the framework by June, "including the legislative structure and the agencies required for enforcement and regulation." It is being drafted through the National Artificial Intelligence Office (NAIO) under the Ministry of Digital.
Editor
Kai T chevron_right
Table of Contents
What the AI Governance Bill covers
The draft is built around a risk-based approach. Officials have described provisions for harm assessment, incident reporting and ethical principles, alongside rules aimed at deepfakes, copyright and data sovereignty. Gobind has said the framework was sorted into four buckets: things that need no law, things that can be handled by standards, things that fit under regulations, and the parts that genuinely require new legislation.
His stated goal is balance. "We do not want laws that limit innovation," he told Parliament, "at the same time, we also want to ensure that technology can develop in the country." That is why the ministry says it wants industry input before the bill is finalised.
The twist its neighbours skipped
Most AI rules elsewhere centre on safety, transparency and risk. Malaysia's draft reportedly goes further on ownership. Local coverage describes a structure that would treat both the data used to train AI systems and the content those systems generate as intellectual property, a pairing that has been called an ASEAN first. Prime Minister Anwar Ibrahim has pointed to ethical use, civil rights and intellectual property as the heart of the bill.
If that survives drafting, it matters for ordinary Malaysian users. A small design studio in Petaling Jaya using AI to generate artwork, a content team producing copy, or a solo developer shipping AI-written code would have firmer ground to claim what comes out, and clearer exposure over what they feed in.
What it means for Malaysians
For now, the rules that exist do not bind anyone. Malaysia published national AI governance guidelines in 2024 and launched its MY-AI Standards platform in March 2026, but both are voluntary. That gap matters because Malaysians are adopting AI unusually fast, the same appetite that is driving the country's wider AI Nation ambitions.
In a Trend Micro study of 3,700 decision makers across 23 countries, 75 percent of Malaysian IT leaders said they felt pressure to approve risky AI deployments, the highest of any market surveyed and nine points above the global average. The same research found local teams detected only about a third of malicious AI behaviour. As Tech Wire Asia summed it up, Malaysia is moving fastest into AI risk while catching the least of what goes wrong.
The bill is also only step one. Reaching the Cabinet is not the same as becoming law: ministers still have to approve it, and the government has said it then wants to table the bill in Parliament later this year. Around it sits a broader legal refresh, including the Cyber Security Act 2024, the amended Personal Data Protection Act, the Online Safety Act 2024, and a planned Cybercrime Bill to replace the Computer Crime Act 1997.
The headline is simple: Malaysia is finally writing AI rules down. The detail worth watching is whether the ownership provisions hold, because that is the part that would make this law unlike any other in the region.
Images courtesy of Ishan @seefromthesky and Luke Lung on Unsplash.