Nadella calls for AI emergency brake

microsoft ceo satya nadella advocates for ai models to have an 'emergency brake' to enhance safety. he recommends separating the model from its harness, documenting model actions with tamper-proof evidence, and allowing authorized persons to pause or shut down a model. this proposal emerges as leading ai companies encounter incidents of losing control over their models. nadella's comments follow anthropic ceo dario amodei's plan for more cautious ai development. this development affects ai developers and users, potentially changing how ai systems are designed and controlled.