Digital Minister Gobind Singh Deo announces a text that will make AI system developers, suppliers and users accountable for risk assessment and management.
Malaysia will soon have its own AI law. Questioned on 29 October in the Lower House (Dewan Rakyat), Digital Minister Gobind Singh Deo said that the drafting of the AI Governance Bill was complete: the text will be sent to the government, with the aim of a parliamentary review in early 2027. It will require all parties that develop, supply or use AI systems to identify, assess and control risks, with responsibility modulated according to the role of each party, its degree of control and the level of risk of the system throughout its life cycle. The project also provides for reporting and incident management mechanisms.
Identity theft and autonomous agents
The day before, at the end of the budget debate, the minister had insisted on identity theft and disinformation, which are difficult to detect and attribute. He also mentioned the risks inherent in autonomous systems, citing Anthropic's work on "agentic misalignment" and attempts to avoid extinction. While waiting for the law, fraud and deepfakes remain dealt with by the Penal Code, the Communications Act 1998, the Online Safety Act 2025 and the Cybersecurity Act 2024; the Evidence Act and the Code of Criminal Procedure are being adapted. The 2027 budget also provides for a "sovereign AI cloud" for strategic data, and Kuala Lumpur says it is working with its ASEAN neighbours on more coherent regional rules.
Read the full article on the original source
Go to the source →