Microsoft Vice Chair and President Brad Smith has called for stronger international rules governing artificial intelligence, arguing that AI systems need clearer safety requirements. In an interview with the BBC published on October 12, 2026, Smith said AI faces fewer safety obligations than everyday consumer products, according to Reuters.
Smith proposed global rules that would require companies to notify people affected by breaches involving rogue AI agents, provide a way to stop a system in an emergency and prepare formal incident reports. Independent evaluators should also be able to assess new AI developments. An AI agent is a system that can carry out tasks with limited human intervention, which makes monitoring its actions important when it operates beyond a simple question-and-answer exchange.
These proposals matter to businesses, students and ordinary users because AI systems are increasingly being used to perform tasks rather than merely generate text. Notification requirements could help affected people understand when a system has acted improperly, while incident reporting could give regulators information about failures. An emergency shutdown mechanism could provide a way for authorised people to interrupt an AI system when necessary. Smith's comments were proposals for regulation, not a report that these requirements had already become law worldwide.
Smith identified Britain's AI Security Institute as well placed to help lead regulatory efforts. Reuters reported that the institute receives access to major AI models before deployment under voluntary agreements so it can assess their capabilities and risks. Britain's AI Minister Kanishka Narayan said the government would consider regulating advanced models if the existing system proved insufficient. The report did not specify a date for new rules to take effect.