OpenAI has urged the UK government to introduce legislation for the companies developing the most powerful artificial intelligence systems, saying ministers should act during a renewed political focus on AI safety. The company said the rules should target advanced AI labs, including OpenAI, rather than startups building less powerful systems.
The intervention followed a report from the cross-party joint committee on human rights calling for a new UK regulatory framework, an independent oversight body and stronger protections against AI-related harms.
What the committee is seeking
The committee, chaired by Labour MP Alex Sobel, said current legislative and regulatory approaches in the UK and elsewhere were not fit for purpose. It cited risks including public face-scanning, explicit deepfakes and disciplinary action against workers without reasonable cause. Related coverage: India’s censorship blocks film on Punjab’s separatist crackdown and human rights abuses.
Sobel said a single AI regulator should set policy, monitor performance and enforce protections across the technology’s supply chain and life cycle. He has also proposed a bill to prohibit the development of superintelligent AI.
Key points
- The human rights committee said higher-risk AI systems should face more demanding obligations.
- It called for some uses of AI that conflict with human rights to be prohibited.
- The committee also sought transparency requirements across the AI life cycle.
- The existing AI Security Institute tests advanced models but cannot compel developers to submit them for scrutiny.
How companies and ministers responded
OpenAI’s head of policy in Europe, Tom Duff Gordon, said the government should take advantage of a political window for legislation. OpenAI chief executive Sam Altman separately said he would welcome a federal framework with consistent safety requirements for frontier AI, while arguing that companies could begin providing confidence without waiting for an antitrust exemption or new legislation.
The call came after Anthropic chief executive Dario Amodei urged leading AI companies and governments to coordinate a global slowdown in development. OpenAI, Google DeepMind and Elon Musk supported the call, while Amodei also sought a US government waiver allowing leading companies to coordinate without anti-competitive-collusion penalties.
A UK government spokesperson said ministers took advanced-AI risks extremely seriously and would consider future measures that were evidence-based and focused on the risks requiring action. Downing Street rejected a call from more than 70 MPs and peers for a ban on creating artificial superintelligence, but said it was exploring targeted interventions addressing the most significant AI-related national security risks.
What happens next
The UK government is considering whether additional targeted measures are needed, while the human rights committee is calling for new legislation and a single independent AI regulator. Read the context: US diplomats stage walkout at UN following French criticism of Trump administration’s human rights record.
No comments yet. Start the discussion.