In a lengthy post on X, Microsoft CEO Satya Nadella outlined his views on the dangers posed by highly advanced AI models and proposed a shift in how the industry approaches safety. Nadella stated that the world can no longer accept AI as a "set of nested black boxes" whose advice and actions are simply accepted or rejected.
Transparency and the Emergency Brake
Nadella called for building a more transparent system where models can be contained, observed, and leave behind "tamper-proof human-readable evidence." He suggested adopting a stance of proactive caution regarding the integrity of these systems.
"We must assume a model is compromised and contain it from the start," he wrote. He likened this approach to an "emergency brake," suggesting that an authorized person should always be able to pause or shut down a model mid-task. According to Nadella, more advanced models will require more advanced containment technologies that the industry needs to standardize.
Industry Alignment and 'Super Intelligence'
Many of Nadella's recommendations align with existing industry calls for timely incident disclosure, independent audits, and verifiable data. However, his emphasis on containment appears to go further than some of his peers. Throughout the post, Nadella also referred to AI as "super intelligence."
