UK AI rules face global pacing debate

UK calls for dedicated AI legislation are intensifying after a parliamentary committee urged an AI bill, an independent oversight body and stronger powers for the UK AI Security Institute. The proposals go beyond the government’s current reliance on sectoral regulators, which lawmakers say leaves fragmented protections, and differ from a separate push to prohibit development of superintelligent AI in the UK.

The government said it takes advanced AI risks seriously and may need to strengthen existing rules as capabilities develop, while keeping any future measures evidence-based and focused on specific risks. Critics and campaigners argue ministers are still too reactive, warning that current systems already raise concerns around cyberattacks, political influence, child safety, discrimination and psychological harms.

The UK debate is unfolding alongside wider calls to slow frontier development. Anthropic chief Dario Amodei has backed “pacing” through regulation targeting all US frontier AI companies, with transparency, third-party auditing and embedded evaluators inside leading labs. OpenAI said it supports stronger UK rules for the most powerful models if startups and smaller developers are not burdened, while human rights officials and researchers warned that industry coordination alone is insufficient without independent verification and aligned international safeguards.