China Tightens AI Safety, U.S. Urged to Reengage
Beijing has made AI safety a national priority with new standards, enforcement, and research; experts want Washington to revive bilateral talks and coordinate on high-risk threats like biological misuse and loss of control.
- China elevated AI safety at the top level, including a Politburo study session and listing AI in its National Emergency Response Plan.
- Regulators now require pre-deployment safety assessments, removed over 3,500 non-compliant AI products, and issued more national AI standards in the first half of this year than in the prior three years combined.
- The volume of Chinese technical papers on frontier AI safety more than doubled over the past year.
- No follow-up U.S. China government meeting on AI risks occurred after May 2024, and bilateral talks have not resumed under the recent administrations.
- Experts propose reviving government channels to share standards, safety evaluations, incident reporting, and emergency response protocols for high-stakes threats such as CBRN and biological misuse.