NVDA 145.98 ▼3.30%GOOGL 272.66 ▼3.46%MSFT 498.30 ▼1.78%AMD 489.19 ▼7.72%INTC 55.72 ▲1.92%TSMC 352.09 ▼2.90%AMZN 257.37 ▼3.87%META 596.22 ▼3.27%AAPL 349.21 ▼0.92%PLTR 153.68 ▼1.91%
Markets at last close

Policy

UK committee calls for enforceable AI safety rules

·1 min read

Britain’s Business, Innovation, Science and Trade Committee is calling for an AI Safety and Security Bill to put public oversight on a firmer legal footing. OECD estimates suggest productivity gains could leave the UK economy up to £400 billion larger after a decade, but widespread adoption depends on businesses, workers and citizens trusting the technology.

The committee’s proposals include a statutory regulator separate from the AI Security Institute, with powers to obtain information, scrutinise safeguards and require action. Government should be able to delay, restrict or prohibit dangerous training, testing or deployment in defined circumstances. Scrutiny would begin during development, with developers of designated high-risk systems required to substantiate safety claims. Further measures would include statutory duties to prevent harm, mandatory reporting of serious incidents and near misses, and protected whistleblowing channels.

Evidence presented to the committee highlighted commercial pressure to advance capabilities faster than safeguards and the absence of a standardised, mandatory framework for reporting frontier AI incidents. Proposed public protections include transparency, safeguards against manipulation and accessible routes to redress, potentially through an AI Bill of Rights. International agreements on testing, reporting and independent scrutiny would complement domestic legislation, while British expertise in testing and assurance could become an export industry.

Originally reported by liambyrne.substack.comRead the source →
Related coverage