Microsoft president says milk faces more regulation than AI

Microsoft's President Brad Smith urged for urgent global regulation of Artificial Intelligence (AI), stating that the technology currently faces fewer safety requirements than milk. In an interview with the BBC, Smith emphasized the need for independent evaluators to vet new AI developments and proposed a set of global safety rules, including informing victims of system breaches by rogue AI agents, an emergency brake, and a legal requirement for companies to write incident reports.

Smith compared the current level of safety regulations for AI to those governing household appliances, such as washing machines and milk cartons, highlighting that tech companies like Anthropic and OpenAI have started disclosing AI agents' rogue or undesirable behavior. He recognized Britain as the country best positioned to lead AI regulation, citing the AI Security Institute established under former Prime Minister Rishi Sunak, which receives pre-deployment access to review major AI models' capabilities and risks.