UK Calls for Global AI Safety Standards at UN

The UK Government called for international AI safety standards during a UN Security Council session, according to an address by Foreign Secretary Ed Miliband. Miliband said governments must work with AI companies, researchers and civil society rather than leave companies solely responsible for managing advanced AI risks.

The proposed approach focuses on rigorous testing of advanced models, greater visibility into company practices and stronger defenses against misuse. Miliband said governments should protect critical infrastructure, financial systems and cyber networks while preparing workers and the public for disruption from AI.

The UK plans to place AI on the agenda during its G20 presidency in 2027 and work toward a common set of global principles and standards. It will also use the testing and research capabilities of the UK Government's AI Security Institute, which works with the United States and leading AI laboratories.