Microsoft’s Satya Nadella says AI models need an ‘emergency brake’
techcrunch ·
Microsoft CEO Satya Nadella has called for improved AI safety measures, emphasizing the need for an 'emergency brake' to control advanced models. In a recent post on X, he argued that organizations must reassess their trust architecture rather than treating Super Intelligence as opaque black boxes. Nadella proposed separating the underlying model from the harness that orchestrates its tasks and externalizing necessary controls and safeguards. He insisted that every significant action taken by a model should be documented with tamper-proof, human-readable evidence. Furthermore, he advocated for systems where authorized personnel can always pause or shut down a model mid-task if issues arise. These comments follow growing industry concerns about incidents where companies have lost control of their AI systems. The statement also comes shortly after Anthropic CEO Dario Amodei published his own plan for more cautious AI development.
techcrunch