Sunak and AI leaders discuss the ‘existential threats’ and fear of misinformation

Rishi Sunak discusses with leading figures in artificial intelligence the need for regulation to mitigate risks ranging from disruption and national security to “existential threats”.

The Prime Minister held a meeting with CEOs of ChatGPT firm OpenAI, Google DeepMind and Anthropic on Wednesday evening to discuss the concerns.

Mr Sunak said AI is the “defining technology of the time” which has the potential to “positively transform humanity”.

But a joint statement from the meeting acknowledged the success of the technology depended on having “the right guardrails” in place so the public can be confident it is safe.

Mr Sunak stressed to Sam Altman of OpenAI, Demis Hassabis of Google and Dario Amodei of Anthropic that regulation must be agile and coordinated internationally.

“The prime minister and CEOs discussed the risks of technology ranging from misinformation and national security to existential threats,” their statement said.

“They discussed safeguards, voluntary actions that labs are considering for risk management, and possible avenues for international cooperation on AI security and regulation.”

Mr. Sunak has advocated the technology’s benefits for national security and the economy, but growing concerns have been raised with the prominence of ChatGPT bots.

Jobs are being put at risk by rapidly developing technology, alongside the concerns discussed by Mr Sunak and the technical figures.

Last week BT ​​Group said it would cut 55,000 jobs by the end of the decade amid plans to shift to AI and automated services.

The former government’s chief scientific adviser, Sir Patrick Vallance, has warned that AI could have an impact on jobs comparable to the industrial revolution.

Earlier this month, Geoffrey Hinton, the man widely seen as the godfather of AI, warned that some of the dangers of AI chatbots are “quite scary” as he quit his job at Google.

Mr Sunak is hardening his tone towards AI. The government’s policy paper on the technology, published less than two months ago, was titled “A pro-innovation approach to AI regulation”.