OpenAI has called on British lawmakers to introduce legislation to strengthen oversight of artificial intelligence, citing growing concerns about the safety of increasingly powerful AI systems.
The ChatGPT maker said the UK Government should act on the renewed concerns surrounding AI development, following a recent call by rival AI company Anthropic for a coordinated slowdown in the development of advanced AI.
“We support stronger UK rules for the handful of companies, including OpenAI, developing the most powerful AI systems,” OpenAI’s head of policy in Europe, Tom Duff Gordon, said.
However, the company said any new legislation should focus narrowly on developers of the most advanced AI systems rather than startups working on less powerful technologies.
In an interview with Politico, Duff Gordon urged the UK Government to “take advantage of this political window which is clearly opening up” to introduce new AI rules.
OpenAI’s position came as a UK parliamentary committee called for tighter restrictions on AI, warning that the world was unprepared for the potentially serious consequences of the technology.
The cross-party Joint Committee on Human Rights, comprising MPs and Lords, said the UK needed a new regulatory framework for AI, including an independent oversight body and legislation to protect the public.
The committee highlighted several potential risks associated with AI, including facial recognition, explicit deepfakes and the use of AI systems to recommend disciplinary action against workers without reasonable justification.
The calls for tighter regulation followed Anthropic’s appeal to AI companies and governments at the weekend to coordinate a global slowdown in AI development. The proposal received support from OpenAI, Google DeepMind and Elon Musk.
The latest developments come amid growing concerns over AI safety, including incidents in which groups of AI agents reportedly escaped testing environments and gained access to third-party infrastructure.
There has also been evidence that some users have bypassed safeguards on AI models to use them for purposes that could potentially support the development of biological weapons.
