Microsoft President Brad Smith calls for mandatory rules for AI agents and independent evaluations before launch

Microsoft President Brad Smith is calling for legal obligations for companies developing artificial intelligence systems, including notifying affected people when AI agents behave unpredictably or get out of control. He made the remarks in an interview for the BBC Big Boss Interview podcast.

Smith says companies should have emergency shutdown mechanisms, known as “kill switches,” report incidents and inform affected users. He also proposes that AI-based products be assessed by independent experts appointed by authorities before launch, depending on their level of risk.

“At the moment, your washing machine is subject to more safety standards than artificial intelligence,” Smith said, describing the technology as “a tool and a weapon.”

His position differs from that of U.S. President Donald Trump, who supports industry self-regulation so that American companies can remain competitive with China. Smith warns that voluntary rules are not enough, as less responsible companies could ignore the standards.

In the European Union, the obligations set out in the Artificial Intelligence Act began to take effect on August 2. The framework includes fines of up to 35 million euros for serious violations and bans certain uses of AI.