Brad Smith says AI has fewer safety rules than a washing machine
🎧 Listen to this story
Microsoft president Brad Smith called for global artificial intelligence rules in an interview with the BBC’s Big Boss Interview podcast. He said that a washing machine and even a carton of milk are subject to more safety standards than the most advanced AI systems, that the technology is advancing “faster than the speed of safety” and that it is both a “tool and a weapon.”
Smith proposed an emergency brake for systems controlling autonomous devices or critical infrastructure, mandatory incident reporting and independent approval of frontier models before launch. According to Xataka, he said the voluntary commitment promoted by President Donald Trump and signed by several companies is necessary but insufficient; Microsoft did not initially sign it, although Smith said it is willing to do so. Xataka also reported, citing Politico, that the White House had allegedly asked Anthropic and OpenAI not to make recent models available to British evaluators; TechCrunch reported that the UK AI Security Institute can still access and test them.
Why it matters · editorial interpretation
Brad Smith’s proposals put international rules, mandatory reporting and independent approval at the center of the debate over AI systems connected to autonomous devices or critical infrastructure. The disagreement over the U.S. voluntary commitment also suggests tension between voluntary measures and formal oversight, although the summary does not indicate how these proposals would be implemented.
Sources
RegulationSecurity