LONDON — Microsoft President Brad Smith called for urgent global regulation of artificial intelligence, including independent evaluators able to vet new developments, warning that AI currently faces fewer safety requirements than milk.

Speaking in an interview with the BBC published on Monday, Smith said a set of global safety rules needed to be in place, including a requirement that a "victim" be informed if their system is breached by a rogue AI agent and that an "emergency brake" be available.

He also said there should be a legal requirement for companies to file "incident reports."

"We know how to do this for other products," Smith said. "We should build on that knowledge and think about how to apply it to AI."

"Right now, your washing machine is subject to more safety rules than AI. Every carton of milk that you see in the grocery store is subject to safety rules and inspection."

Tech companies such as Anthropic and OpenAI have started disclosing rogue or undesired behavior by AI agents, part of a global debate over the fast-advancing technology that has included researchers' warnings of an eventual existential threat to humanity. Microsoft is itself one of OpenAI's most significant backers, with a multibillion-dollar investment and technology partnership that embeds the startup's models across its products.

Some companies have called for AI development to slow down, but that has prompted debate over whether executives have ulterior motives, such as setting rules that preserve the lead of their firms, pre-empting a broader government backlash, or simply slowing breakneck spending.

Smith said Britain was the country best placed to lead the way on AI regulation, citing the AI Security Institute, which was launched under former Prime Minister Rishi Sunak at the 2023 Bletchley Park AI Safety Summit. He added that officials in Washington viewed London very differently from Brussels, which regulates AI according to risk level under its AI Act, the world's first comprehensive AI law whose obligations are being phased in.

Britain's institute receives pre-deployment access to major AI models under voluntary agreements, allowing it to assess their capabilities and risks. Britain has so far overseen AI through existing regulators applying non-statutory principles rather than a dedicated AI law.

AI Minister Kanishka Narayan has told Reuters that Britain would consider regulating advanced models if the current system proved insufficient.

— Reuters