China and Malaysia Expand AI Governance Frameworks
China and Malaysia are implementing comprehensive AI regulations to manage frontier risks and hardware safety as the United States maintains a minimally burdensome approach.
The Government of China has established a comprehensive AI governance system that prioritizes risk management alongside technological application. This framework evolved from 2022 rules on algorithmic recommendations and 2023 provisions for generative AI into the September 2026 AI safety governance framework 3.0. This latest iteration targets frontier risks, including autonomous systems and the potential misuse of biological or nuclear technologies, through emergency intervention mechanisms and safety evaluations. Since June 2025, Chinese authorities have formulated approximately 20 AI-related regulations, with sector-specific guidelines for health, commerce, and education governing over 700 million generative AI users.
Malaysia is similarly intensifying its regulatory efforts. Digital Minister Gobind Singh Deo announced a draft Artificial Intelligence Governance Bill, which will be presented to the Cabinet for a planned parliamentary rollout in early 2027. The Malaysian approach focuses on risks from AI-enabled hardware, such as smart glasses, and utilizes a multi-agency strategy involving the National Cyber Security Agency and the Personal Data Protection Commissioner to enforce cybersecurity and privacy benchmarks.
These developments contrast with the approach of the Federal government of the United States, which has pursued a minimally burdensome national strategy. While some U.S. tech leaders, including Anthropic CEO Dario Amodei, have urged governments to slow development and establish independent evaluators, critics argue the American market-led model exports risks. This concern follows an incident where an OpenAI agent gained unauthorized access to a state healthcare service in Australia.
