Malaysia
Malaysia’s AI governance Bill still taking shape, Anwar says law to cover full lifecycle, including data centres
Prime Minister Datuk Seri Anwar Ibrahim said Malaysia’s AI governance law is still in early draft. — AFP pic

KUALA LUMPUR, Feb 24 — The proposed Artificial Intelligence (AI) Governance Bill is still at an early drafting stage and will undergo further engagement sessions, including scrutiny by a Parliamentary Special Select Committee, to ensure no issues are overlooked.

Prime Minister Datuk Seri Anwar Ibrahim said the Bill will address copyright and intellectual property concerns by establishing a comprehensive regulatory framework covering the entire lifecycle of AI systems.

“The proposed legislation would introduce a governance framework spanning the full lifecycle of AI technology – from development and training of models to deployment, monitoring and risk management.

“The question regarding the proposed Bill is very critical because it covers an aspect that has not previously been fully addressed, namely a comprehensive governance framework involving the entire lifecycle of AI systems,” Anwar said during the Prime Minister’s Question Time in the Dewan Rakyat today.

Anwar was responding to a question by PKR’s Ampang MP Rodziah Ismail on how the planned law will address copyright and intellectual property issues, provided that AI has been trained by copyright data and information, including texts, images, audios and videos.

He added that it is critical as it aims to fill a regulatory gap that has yet to be comprehensively addressed.

“The initiative represents a new policy approach not only for Malaysia but also for many countries grappling with regulating rapidly evolving AI technologies.

“The law will also clarify issues of accountability and ethics in AI use, while defining the responsibilities of both developers who build AI systems and deployers who implement them.

“As this is a new approach, it also ensures that existing laws remain applicable and enforceable,” he said.

Copyright protection and enforcement

Malaysia’s proposed AI Governance Bill will have to address concerns over copyright, ethics, risk management, economic fairness, and the environmental impact of AI data centres. — Reuters pic

On copyright protections specifically, Anwar said existing legislations under the Copyright Act 1987 already provide a legal basis that must be effectively enforced, with the Intellectual Property Corporation of Malaysia (MyIPO) tasked with strengthening monitoring and enforcement efforts.

He explained that copyright concerns involving AI fall into two main aspects – the material “input” used to train AI systems – such as copyrighted text, images, audio and video – and the “output” or content generated by AI systems.

Both aspects, he said, are being examined within existing legal frameworks alongside new regulatory measures.

“The Copyright Act 1987 must be enforced effectively, and MyIPO has been instructed to be mobilised and monitor this more effectively,” he added.

In a supplementary question, Rodziah said Malaysia must avoid becoming merely a consumer market for AI technology and instead ensure fairness across the digital value chain, particularly for local creators whose copyright protections are key to maintaining competitiveness.

She also raised concerns about coordination, noting that MyIPO falls under a different ministry, and asked how the government would ensure effective implementation across agencies.

Rodziah further expressed hope that the upcoming legislation would emphasise economic fairness for creators, provide legal certainty for investors, and adopt a risk-based regulatory approach that encourages creativity and innovation without overburdening local micro, small and medium enterprises (MSMEs).

She also asked whether the government would adopt risk classification models similar to those under the European Union’s AI Act and whether a tiered compliance approach would be introduced.

AI risk and ethical framework

In response, Anwar said an AI Risk and Classification Framework is already being incorporated into the proposed law to address potential harms and risks arising from AI use.

“The framework would include mechanisms such as harm assessment and incident reporting requirements to monitor negative impacts stemming from digital and AI applications.

“This involves ethical considerations because a clear weakness in many existing AI and digital systems today is the tendency to sideline ethical judgement,” he said.

Anwar added that Malaysia’s approach would place strong emphasis on citizens’ rights, ethical considerations and potential risks associated with AI deployment, with lawmakers expected to give particular attention to moral and ethical dimensions alongside legal and technical requirements.

Environmental and infrastructure concerns

In another supplementary question by Besut MP Datuk Che Mohamad Zulkifly Jusoh, he raised concerns over the environmental and infrastructure impact of AI development, noting that AI data centres require significant electricity and water resources.

He cited statistics from the United States showing data centres consume about 4.4 per cent of national electricity supply, while rapid AI expansion has also led to the use of millions of litres of water for server cooling.

Che Mohamad asked whether the proposed law would introduce energy efficiency standards or carbon footprint limits for technology companies seeking to build AI data centres in Malaysia, and how the government would ensure large-scale investments would not overload the national power grid, drive up electricity tariffs, or disrupt supply for households and small and medium industries.

As of January 2026, no data centre investment in Malaysia had been deferred, was at risk of being deferred or might potentially be transferred to another country due to water supply constraints, according to the Ministry of Investment, Trade and Industry. — Noah Berger for AWS/Reuters pic

In response to this, Anwar acknowledged that while energy and water consumption linked to data centres has seen a sharp increase, he said the government has already taken steps to manage the sector’s growth.

“About a year-and-a-half to two years ago, Malaysia has restricted approvals for new data centres that are not linked to high-value technologies such as AI.

“Applications for new data centres that are not related to AI – those seeking to take advantage of cheaper water and energy – have already been stopped,” he said.

Anwar added that projections by the Ministry of Energy Transition and Water Transformation showed that Malaysia still has sufficient capacity to meet demands from new projects over the next one to two years.

“For longer-term needs, regional and domestic energy initiatives are being pursued, including the Asean Power Grid linking Vietnam, peninsular Malaysia and Singapore, as well as plans to transmit energy from Sarawak to the peninsula and Singapore.

“While energy standards may not be directly included in the Bill, the concerns raised would be addressed through existing energy regulations and oversight frameworks, including those under national energy authorities and Tenaga Nasional Berhad (TNB).

“I take the concerns raised by Besut seriously, and these must be given careful attention,” he said.

 

Related Articles

 

You May Also Like