ع
Digital Minister Gobind Singh Deo informed the Dewan Rakyat that high-risk artificial intelligence systems will encounter stricter safety controls, mandatory testing, monitoring and human oversight under the forthcoming AI Governance Bill. The minister described a risk-based approach that evaluates controls according to the potential harm an AI system may cause rather than the industry in which it operates, ensuring direct impacts on safety, fundamental rights and individual well-being receive appropriate regulation. Gobind Singh Deo added that the legislation will function as a cross-sectoral framework designed to complement rather than replace existing laws while prioritising safety alongside technological advancement. The government recognises the significant effects of AI on the economy, society and national administration, prompting continued focus on risk mitigation as capabilities expand.
The bill adopts a three-tier risk model with unacceptable risk applications prohibited in cases involving intent to harm, high-risk systems requiring assessments, documentation and oversight, and low-risk uses subject to lighter obligations, according to details shared in recent parliamentary sessions. This structure enables proportionate responses that strengthen prevention and accountability across the entire AI lifecycle from development through deployment and operation. Gobind Singh Deo had previously noted in June that the framework establishes clearer responsibilities for developers, providers, users and operators while allowing enforcement under relevant statutes for any unlawful acts. Such measures address risks more comprehensively by combining early prevention with targeted enforcement against misuse.
Gobind Singh Deo stated on September 10, 2026 that the draft bill had been completed after the Digital Ministry conducted workshops and engagement sessions to incorporate stakeholder views and was then advancing to the cabinet for consideration. He expressed hope that tabling could occur in the current quarter or by the first quarter of next year at the latest, building on July consultations run by the National AI Office to refine the governance framework. The process reflects ongoing efforts to introduce structured oversight that forms the basis for responsible AI adoption nationwide. These steps align with broader initiatives to enhance understanding of AI risks and opportunities among public and private entities.
AI governance encompasses standards, regulations and legislation, with the ministry already advancing elements addressable through standards via the MY-AI Standards platform developed in partnership with Sirim Bhd. Gobind Singh Deo has clarified that this multi-pronged strategy permits quicker implementation of technical benchmarks while the bill undergoes legislative review to provide enforceable mechanisms. The approach avoids siloed responses by promoting uniform technology risk management across sectors, as outlined in multiple ministry briefings throughout the year. Collaboration with industry and regulatory bodies has shaped these standards to support practical compliance without stifling innovation.
An August 2026 study commissioned by Amazon Web Services found that 38 percent of Malaysian businesses, or more than 3.4 million entities, now use at least one AI tool, marking an increase from 27 percent the previous year with nearly one million new adopters. The Strand Partners survey highlighted that financial services and manufacturing lead adoption at 53 percent and 50 percent respectively, although two-thirds of users remain focused on basic applications such as chatbots or off-the-shelf productivity tools. Only 19 percent of businesses possess a formal strategy to scale AI across functions, underscoring gaps in maturity despite rapid uptake. AI Malaysia Berhad, established in July 2026 as the permanent successor to the National AI Office, coordinates these developments through the National AI Action Plan 2026-2030 that integrates governance with talent development, infrastructure and sectoral applications.
The proposed legislation includes mechanisms for reporting AI-related incidents, particularly sophisticated deepfakes, to enable faster identification of threat patterns and prompt mitigation before widespread harm occurs, according to earlier statements by Gobind Singh Deo. This builds toward a Central AI Authority that would oversee implementation and enforcement while maintaining alignment with international best practices amid global regulatory momentum. The bill’s emphasis on accountability ensures entities involved in AI systems take reasonable measures scaled to risk levels, supporting Malaysia’s objective of becoming a responsible AI hub by 2030. Parliamentary discussions continue to refine these provisions based on ongoing stakeholder feedback.
