Satya Nadella warns companies using proprietary AI models

Microsoft CEO Satya Nadella has issued a sharp warning to enterprises relying on proprietary AI models: you are paying for intelligence twice, once with money and again with the proprietary knowledge you must reveal to make the model useful.

He argues that as companies feed prompts, corrections, and usage data into models from labs like OpenAI and Anthropic, they are essentially training their future competitors.

This is not a fringe concern—it echoes warnings from prominent VCs and tech leaders, and now carries the weight of a CEO whose company is deeply invested in the very labs he warns against.

The core tension is that model makers claim fair use rights to train on public data while simultaneously locking down distillation, the process by which enterprises could study and replicate those models for themselves.