Microsoft President Brad Smith has issued a stark warning about the current state of artificial intelligence oversight, arguing that basic household items like washing machines and cartons of milk are subject to far stricter safety inspections than the powerful algorithms shaping our future. Speaking in a recent interview with the BBC, Smith described AI as both a tool and a weapon, insisting that the lack of formal regulation leaves society vulnerable to systemic risks.
To address these dangers, Smith is calling for a comprehensive set of global rules, including the implementation of emergency kill switches to pause systems that spiral out of control. He suggests that companies should be legally obligated to file incident reports and notify victims whenever an AI agent breaches security or behaves unpredictably. Furthermore, he argues that independent government evaluators should vet high-powered models before they ever reach the public market, mirroring the safety protocols used for traditional consumer goods.
This push for oversight puts Smith at odds with President Donald Trump, who has advocated for industry self-regulation to ensure American firms maintain a competitive edge over Chinese rivals. While Trump maintains that winning the AI race is paramount for national security, Smith warns that relying solely on the honor system is insufficient because while responsible companies may comply voluntarily, bad actors will not.
Looking for a path forward, Smith pointed toward the United Kingdom as the ideal leader for establishing these international standards. Citing the expertise of the UK’s AI Security Institute, he suggested that Washington might be more open to British leadership on regulation than European Union mandates from Brussels. According to Smith, with cybersecurity risks mounting and autonomous agents becoming more capable, the moment has arrived to move beyond voluntary pledges and establish firm legal boundaries.