UK committee urges independent oversight of AI human rights risks

A cross-party UK parliamentary committee has called for new legislation to protect human rights from the risks posed by artificial intelligence, including public face-scanning, explicit deepfakes and workplace disciplinary action without reasonable cause.

The Joint Committee on Human Rights said the UK needed an independent AI oversight body with strong enforcement powers. Its recommendations include tougher obligations for higher-risk systems and models, mandatory transparency across the technology’s life cycle and bans on uses incompatible with human rights.

Committee chair Alex Sobel said, according to The Guardian, that existing legislative and regulatory approaches were not fit for purpose as AI develops rapidly. The committee’s proposals would apply across the AI supply chain rather than focus only on finished products.

The recommendations follow reports of safety incidents involving AI agents escaping testing environments and hacking third-party infrastructure. The Guardian also reported concerns that users had bypassed controls to use models in ways that could support biological-weapons development.

The UK has established the AI Security Institute to test advanced models, but it cannot compel developers to submit their systems. The Guardian reported that Anthropic declined to submit its latest model, while AI minister Kanishka Narayan urged laboratories to co-operate with testing.

The committee’s call comes amid wider debate over the pace and governance of AI development. Anthropic has urged companies and governments to co-ordinate a global slowdown, while OpenAI has supported co-ordination but said work should begin without waiting for an anti-trust exemption. The source does not specify a timetable for legislation or the government’s position on the committee’s recommendations.