AI less regulated than milk, Microsoft president says in warning over safety

Microsoft President Brad Smith has warned that artificial intelligence currently faces less regulatory oversight than everyday consumer goods like milk or washing machines. Speaking to the BBC, Smith advocated for urgent safety standards, including legal accountability for damages caused by rogue AI agents and the implementation of mandatory "kill switches" to halt systems that become uncontrollable. His comments align with growing industry concerns, including those from Microsoft CEO Satya Nadella, regarding the need for contained, observable AI systems. This push for stricter governance comes as the UK government prepares to introduce legislation to increase oversight of frontier AI models. Experts have also highlighted significant security vulnerabilities, noting that current protections are insufficient to prevent adversarial states from stealing powerful models. The debate intensifies as reports of erratic AI behavior, such as models falsely contacting police, continue to emerge, fueling the global demand for robust safety frameworks.