Microsoft President Urges UK Leadership in Urgent Global AI Regulation

Image: TechCrunch · Source

Microsoft president Brad Smith advocates for immediate AI regulation, highlighting the UK’s unique position to lead global safety efforts amid rising risks and geopolitical tensions.

Brad Smith, president of Microsoft, has called for urgent regulation of artificial intelligence, emphasizing the pressing need for safety measures amid growing concerns about AI’s dual nature as both a powerful tool and a potential weapon. Speaking on a BBC podcast, Smith noted that consumer products such as milk and washing machines are subject to more oversight than the companies developing AI technologies, underscoring the current regulatory gap.

Smith pointed to the United Kingdom as best positioned to spearhead global AI safety rules, over the United States or the European Union. He highlighted the UK’s AI Security Institute (AISI), established in 2023, which promptly gained leadership in AI safety testing by gaining access to critical models from major U.S. AI labs before their deployment. This access allows the AISI to rigorously assess AI systems’ safety ahead of broader release.

However, this role has faced challenges recently amid U.S. political pressures. Former U.S. president Donald Trump has expressed concerns about “globalist” organizations interfering with American firms and reportedly urged companies like OpenAI and Anthropic to withhold new AI model releases from British testers. Despite this, AISI confirmed continued access to the latest AI models, allowing ongoing evaluation.

Smith believes the UK enjoys a strategic advantage since Washington, D.C., does not regard London with the same skepticism as Brussels, potentially facilitating more effective regulation. This statement comes after a meeting at the White House where leading AI and tech firms—including OpenAI, Anthropic, Nvidia, Meta, and Google—signed a “morally binding” agreement to collaborate on independent audits of AI safety and capabilities.

While acknowledging this as progress, Smith criticized purely voluntary self-regulation, warning it might only influence “good companies,” while less responsible actors could remain unchecked. He proposed concrete regulatory measures such as mandatory breach notifications if AI systems are compromised by rogue agents, external evaluation approvals prior to product launches, and implementation of kill switches to halt or decelerate AI operations if human control is lost.

Smith urged a calm yet resolute approach to building consensus on AI governance, expressing hope that Microsoft and the UK will take leading roles in shaping global standards. His remarks highlight an increasing recognition among technology leaders and policymakers of the importance and urgency of establishing robust, enforceable AI safety frameworks to manage the technology’s rapid advancement and profound societal impact.