AI is less regulated than milk, says Microsoft president

Microsoft president Brad Smith went on the BBC and made a blunt comparison: AI, he said, is regulated less tightly than a carton of milk or a washing machine. That’s not a throwaway line. It was his way of arguing that the industry has moved faster than the rules built around it, and that the gap is now wide enough to matter.

Smith wants global safety rules for AI, not just vague promises from companies. He called for emergency brakes — the industry’s kill switches — so systems can be paused or slowed if they start to go off the rails. He also argued for a legal duty to file incident reports and to tell victims when a rogue AI agent breaches their system.

And he went further. Smith said outside evaluators, including independent government assessors, should sign off on some AI products before release. His point was simple: other technologies already face this kind of scrutiny, so AI can too. He said firms should build on what regulators already know rather than pretending this is a totally new problem with no precedents.

The timing puts him squarely against Donald Trump, who wants the sector to regulate itself so US companies do not fall behind Chinese rivals. Smith, by contrast, said AI is both a tool and a weapon, and that the moment to start setting rules is now. He argued that the UK is unusually well placed to lead because of its AI Security Institute, and said Washington may be more open to UK-made standards than to those from Brussels.