Microsoft President Brad Smith calls for global AI safety rules

Microsoft president Brad Smith has called for urgent global regulation of artificial intelligence. According to Cyprus Mail, Smith said AI systems currently face fewer safety requirements than a carton of milk or a household appliance, a gap he wants closed with binding rules.

Smith made the case in an interview with the BBC published on Monday. He argued that a single set of international safety rules is needed rather than a patchwork of national approaches.

His specific proposals include notifying affected parties when a system is compromised by an uncontrolled AI agent, and putting an emergency shutdown mechanism in place so that a risky deployment can be cut off before harm spreads.

He also backed a legal obligation for companies to prepare and file AI incident reports. Safety rules and inspections already apply to many other products, he noted, and that experience should be carried over to artificial intelligence.

On where regulation should be anchored, Smith pointed to the United Kingdom, saying it has good foundations for leadership thanks to the AI Safety Institute established under former prime minister Rishi Sunak. Under voluntary agreements with large-model developers, the institute gains pre-deployment access to assess capabilities and risks.

According to the report, technology companies including Anthropic and OpenAI have already begun reporting uncontrolled or unwanted behaviour by AI agents, giving the voluntary-plus-evaluation approach real cases to work with.

UK AI minister Kanishka Narayan told Reuters that the country would consider regulating advanced models if the current approach proves insufficient, which leaves the voluntary framework open to becoming mandatory.

For a platform company like Microsoft, pushing for global rules is both a position and an interest: clear rules make the compliance cost of launching, testing and cross-border deployment more predictable, while fragmented national regulation raises it.

The question now is whether Smith's call feeds into concrete legislation, and how Britain's choice between voluntary agreements and mandatory oversight shapes the pace elsewhere.