Microsoft president Brad Smith has called for global safety rules for artificial intelligence, arguing that AI is currently subject to fewer safeguards than everyday products such as washing machines and milk.
Smith proposed emergency mechanisms that could pause or slow an AI system if it moved beyond human control. He also backed legal requirements for companies to file incident reports and notify people affected when an AI system was breached or behaved dangerously. In addition, he said independent government evaluators should determine when a product is safe enough to be released.
His position contrasts with Donald Trump’s support for rapid US development and industry self-regulation. Trump has argued that the United States must move quickly to avoid losing ground to China, while Smith said voluntary commitments are not sufficient because responsible companies are more likely to follow them than less responsible firms.
Smith pointed to the United Kingdom as a strong candidate to lead international work on AI standards.
The dispute follows a White House meeting at which leaders from OpenAI, Anthropic, Nvidia, SpaceX, Meta and Google agreed to work with independent auditors on whether their systems function as intended and avoid unintended access to technical systems. Smith said Microsoft supported that commitment but considered it only a necessary first step, arguing that binding rules are needed to address companies that do not adopt strong safeguards voluntarily.