He stressed that the prime minister’s planned summit on the future of technology would bring together “like-minded” countries.
When asked about the biggest dangers of AI when attending the Commons Liaison Committee, Sunak emphasized “massive social change”, “abuse” risks and “national security risks”.
He said the “existential risks” of AI need to be taken seriously because there is “a lack of understanding of the potential of these models at this time.”
The prime minister spoke of the benefits of AI, which could tackle “incurable diseases such as cancer and dementia, as well as new ways of growing crops.”
He also referred to a recent paper that estimated that GDP could grow by 10% over the next decade and productivity could improve.
Sunak was asked if the autumn conference would be open to all countries as well as the International Atomic Energy Agency (IAEA), including China and Russia.
“AI doesn’t respect national borders, so it brings together people who are thinking about these things in similar ways to exchange ideas and share information,” he said.
Sunak added: “In conversations with the companies themselves, we would all benefit from listening and speaking to each other.
“We are far from establishing an IAEA equivalent for AI, they are a long way off.
“But first and foremost, it would be wise to discuss the matter with like-minded countries.”
The prime minister declined to answer questions about whether the AI bill would be included in the king’s speech.
“What we have to do, and I think we can probably do a lot without the law, is sit down and think about what safety features and guardrails we want to put in place,” he said.
Mr Sunak said the UK “can lead” in AI safety research “because it already has great capabilities”.
