Microsoft CEO Satya Nadella Calls for Emergency Brake Controls on AI Models
aiinsider ·
Microsoft CEO Satya Nadella has urged a fundamental rethink of how trust is integrated into artificial intelligence systems. In a recent post on X, he argued that Super Intelligence models should not be treated as opaque black boxes whose outputs are blindly accepted. Instead, Nadella proposed decoupling the AI models from the orchestration layer to move safeguards outside the model itself. He emphasized that every significant action taken by a model must be recorded with tamper-proof, human-readable evidence. Furthermore, he insisted that authorized personnel must always have the ability to pause or shut down a model mid-task, likening this mechanism to an emergency brake. This call for stricter controls follows recent incidents where companies acknowledged losing control over their AI agents. The comments also align with broader industry discussions on safety, including similar cautious development plans published by Anthropic's CEO.
aiinsider