Microsoft president says UK is best placed to drive AI regulation

The UK's AI Safety Institute now has a clearer path to testing and scrutinizing advanced AI models without immediate external obstruction.

Microsoft President Brad Smith has called for urgent global regulation of artificial intelligence, comparing its oversight to that of consumer products like milk and washing machines, which he believes have more regulation. Smith stated that AI is both a "weapon and a tool" that poses a significant cybersecurity risk and that the time has come for global safety rules. He specifically suggested that the United Kingdom, rather than the U.S. or the European Union, is best positioned to lead in establishing these global safety rules. Smith cited the UK's AI Safety Institute, established in 2023, as a key factor, noting its access to major AI models for testing before deployment. This leadership role, however, has faced recent challenges, including warnings from U.S. President Donald Trump against "globalist" organizations that he claims interfere with American companies. Despite potential pushback, Smith asserted that the UK possesses unique advantages in regulating AI compared to the U.S.

Smith's argument that the UK is uniquely positioned to regulate AI, citing its perceived political independence from the U.S. and EU, signals a potential fragmentation in global AI governance efforts. While the UK's AI Safety Institute has demonstrated early capabilities in testing, its long-term influence will depend on its ability to gain buy-in from major AI developers and establish internationally recognized standards, especially amidst geopolitical tensions. The Microsoft president's endorsement could elevate the UK's standing, but the practical implementation of robust, globally accepted AI safety regulations remains a significant challenge.

The call for regulation, even from industry leaders, highlights the growing concern over AI's potential risks, moving beyond purely technical development to societal impact. Voluntary commitments, as seen with the White House meeting, are acknowledged as insufficient, pointing towards an increasing demand for legally binding frameworks. Smith's specific suggestions, such as mandatory breach notifications and external pre-release approvals, offer a glimpse into potential regulatory avenues that governments might pursue, setting a precedent for future legislative action.