NEW YORK – Burnham: AI safety should be governments’ responsibility

NEW YORK, Sept. 23 /ATSH-DPA/ — British Prime Minister Andy Burnham said governments must set common rules for the development of artificial intelligence (AI), particularly for the most advanced systems.

Burnham said Britain wants to broker an international agreement with a single set of principles and standards to make AI development safer and more transparent.

He said AI could bring major benefits in health care, the economy, education and defense, but warned the technology could cause job losses and serious problems if not properly controlled.

Burnham stressed that key decisions about AI should not be left to companies alone, but should be taken by elected governments.

Clash with the US

Burnham’s stance contrasted with that of US President Donald Trump.

He told the UN that the United States opposes efforts to impose a global control system on AI.

Trump said the US would continue to develop what he called “superintelligence,” arguing it should be encouraged rather than restricted.

What does Britain want?

Britain aims, during its G20 presidency, to push for an agreement among countries on: common rules for AI; greater transparency from companies; safety checks for new AI models; reporting of serious incidents, including situations where AI behaves deceptively or humans lose control of it.

Meanwhile, major tech companies, including Meta, Google, OpenAI and Anthropic, have been summoned by British lawmakers to discuss whether AI models should undergo independent safety tests before being put into use.