Nadella's 'Reverse Information Paradox': Enterprises 'Pay for Intelligence Twice' — and OpenAI/Anthropic Are Being Hypocrites on Distillation

TechCrunch·high signal

In a July 13 blog/X essay, Microsoft CEO Satya Nadella argued that companies using closed frontier models pay once in tokens and again by leaking proprietary knowledge — the prompts, tool traces, and especially corrections that get 'distilled into institutional know-how' by the provider. He called it 'ironic' that OpenAI and Anthropic train on public data under fair use and learn from customer 'exhaust' while banning distillation of their own models, and floated a new AI-era patent concept to let firms guard IP. The framing is a direct competitive shot positioning Microsoft/open models against the closed labs.