Microsoft president calls for mandatory AI safety standards and incident reporting

Microsoft President Brad Smith called for mandatory global AI safety standards, including emergency controls to stop systems that spiral out of control, The News International reported. His proposed requirements include incident reporting, notification of victims when systems are breached or acted on by rogue AI agents, and independent evaluation before new AI products are released. Smith praised a recent White House pledge under which major technology companies agreed to independent audits, but argued that voluntary safeguards are no longer sufficient. He said less responsible companies are the reason legal requirements are needed.