Microsoft president urges global AI regulation and legal duties for rogue systems

Brad Smith, the president of Microsoft, has argued that artificial intelligence currently faces less regulatory scrutiny than everyday consumer products such as milk and washing machines. Speaking on the BBC’s Big Boss Interview podcast, Smith called for the implementation of global safety rules to govern the sector. He stated that there is an urgent need for a legal obligation requiring companies to notify victims if their systems are breached by a rogue AI agent. Smith also advocated for a framework in which outside evaluators sign off on AI products before they are released to the public.

His remarks place him at odds with the position of US President Donald Trump, who has argued that the sector should regulate itself to prevent American firms from losing their competitive edge to Chinese rivals. Smith countered this view by drawing a direct comparison to existing consumer safety standards. He noted that every carton of milk in a grocery store is subject to strict safety rules and inspection, whereas AI systems currently lack comparable oversight. He described AI as both a tool and a weapon that requires immediate regulatory attention to ensure public safety.

Smith proposed several specific measures to address these risks. He said firms need emergency brakes, often referred to in the industry as kill switches, to pause or slow down AI if it gets out of control. He further suggested that there should be a legal requirement for companies to write incident reports and inform victims when technology goes rogue. Additionally, he recommended that independent government evaluators should vet new AI technology to ensure it meets safety standards. Smith argued that society already knows how to regulate other products and should apply that existing knowledge to the AI sector.

The Microsoft executive praised the United Kingdom as being best placed to drive forward these global rules. He cited the AI Security Institute, which was launched during the government of former prime minister Rishi Sunak, as evidence of the country’s technical capability. Smith stated that most people in the tech sector would identify the UK as the leader in this area. He suggested that the US government would be more accepting of UK-led standards than those proposed by the European Union, noting that Washington does not view London in the same negative light as it views Brussels.

Smith acknowledged the recent efforts by the US administration to engage with the industry. At the end of the previous month, President Trump hosted leaders from OpenAI, Anthropic, Nvidia, SpaceX, Meta, and Google at the White House. These executives signed a morally binding document agreeing to work with independent auditors to assess their AI systems. Smith said Microsoft was happy to sign the statement and considered it a good and necessary step. However, he maintained that it was not sufficient on its own.

He warned that self-regulatory regimes face a significant challenge because responsible companies will generally comply with guidelines. The real concern, Smith argued, lies with companies that are not as responsible. He called for a calm but firm approach to bringing people together, expressing hope that Microsoft and the UK could play a leading role in establishing these international standards. The discussion highlights the growing divide between industry leaders who seek robust regulation and political figures who prioritise rapid development and national competitiveness.