OpenAI has called for Britain to legislate specifically for frontier AI as a parliamentary committee warns that the country’s existing regulatory system is not fit for purpose.
The company says stronger requirements should apply to the small number of developers building the world’s most powerful AI systems.
Parliament’s cross-party Joint Committee on Human Rights is going further.
It wants a new AI Bill, mandatory transparency requirements and a single independent statutory regulator with stronger powers over higher-risk systems.
Committee chair Alex Sobel says no existing regulatory approach anywhere in the world is “fit for purpose”.
Britain already has an institution specifically tasked with examining advanced AI.
There is an obvious limitation.
The AI Security Institute cannot compel developers to submit frontier models for testing.
Anthropic declined to provide its latest model to the institute last week.
That does not establish anything about the safety of the model.
It does demonstrate what voluntary oversight means in practice.
Regulation with teeth
A safety institute that depends on the cooperation of the companies it supervises has limited power when that cooperation disappears.
Giving regulators compulsory testing powers would solve part of that problem.
It would create another.
A government capable of deciding which AI systems can be tested, restricted or potentially prevented from deployment acquires considerable influence over one of the most important technologies being developed.
There are competition consequences too. Compliance regimes that the largest laboratories can absorb may become formidable barriers for smaller challengers.
Britain therefore faces two risks in opposite directions.
A regulator without meaningful powers may be unable to supervise frontier AI.
A regulator with badly designed powers could become a gatekeeper for innovation, competition and potentially speech.
Voluntary oversight has an obvious weakness.
The difficult part is deciding what replaces it.
Sources
- The Guardian: OpenAI urges UK lawmakers to strengthen frontier-AI regulation
- UK Parliament: Human Rights and the Regulation of AI inquiry
Join the Dissenting Citizen Newsletter
Independent news and commentary delivered directly to your inbox.
Breaking stories, analysis and commentary throughout the day.
Follow @VoxDissent →