Satya Nadella says we should assume all AI models are ‘compromised’

What's happening

Microsoft CEO Satya Nadella warned in a lengthy post on X that we should assume all AI models are "compromised" and that AI can no longer be treated as "a set of nested black boxes" whose advice and actions we simply accept. Nadella said it's time "to step back and assess the trust architecture" of AI and argued models need an "emergency brake." Separately, Microsoft president Brad Smith said AI today has less regulation than milk or washing machines and called for urgent oversight, naming the UK as well placed to drive AI regulation.

Why it's trending

Leaders at Microsoft are publicly demanding immediate trust, safety and regulatory fixes as AI capability and deployment accelerate.

SignalHolding at its usual pace, confirmed across 3 independent source types.