Microsoft's CEO warns companies are paying for AI twice, once in cash and once in secrets
Satya Nadella says every prompt, correction and workflow businesses feed into AI models is quietly training the companies they pay. He wants enterprises to take back control.
Key points
- Microsoft CEO Satya Nadella published a blog post on Monday warning that enterprises unknowingly hand over proprietary business knowledge every time they use a third-party AI model.
- Nadella argues that the prompts and corrections companies feed into AI systems become training data that makes the model smarter, potentially at the customer's expense.
- Open models running on a company's own servers accounted for 29% of all traffic through Vercel's AI routing service last month, suggesting a real shift is already underway.
- In February 2025, Anthropic accused Chinese open-source AI developers of sending millions of prompts to its Claude model to copy its behaviour into cheaper alternatives.
Satya Nadella runs Microsoft, the company that has poured billions into OpenAI and holds a stake in Anthropic. So when he tells enterprises to be careful trusting the very AI labs he's backed, people pay attention.
In a blog post published Monday, Nadella named a worry that's been circulating quietly in boardrooms. Companies pay for AI tools in money, he says. They also pay in something they rarely notice: the specific, hard-won knowledge of how their business actually works.
Every time a staff member types a prompt into an AI tool, every time they correct a wrong answer, the model learns something. It learns your terminology, your edge cases, your standards. Nadella calls this knowledge the "exhaust" of AI use. It's also, he argues, some of the most valuable information a company owns.
"Every correction is distilled into institutional know-how," he writes. "The kind of knowledge a competitor could never buy."
His concern is that the terms AI providers attach to their services often let them learn from that exhaust. A company feeding its internal processes into an AI model could, in effect, be training a future competitor.
Should businesses be worried about this?
Yes, but the risk is practical, not theoretical. AI providers differ in what their terms allow. Some contractually promise not to train on your data; others reserve the right to do so unless you opt out. Most businesses have never read those clauses.
Nadella also raises a fairness point. AI labs argue they have a legal right to train on publicly available internet data. He finds it inconsistent that the same labs then forbid customers from "distillation", which means using a model's own outputs to train a smaller, cheaper model. Anthropic made headlines in February by accusing Chinese AI developers of doing exactly that to its Claude model and calling for tighter government export controls.
His proposed fix is predictable for a cloud computing CEO. He wants enterprises to keep ownership of their prompts and feedback, build systems that can swap between AI providers easily, and run sensitive workloads on private infrastructure. That infrastructure, conveniently, often means Microsoft Azure.
But the market is already moving this way. Idit Levine, founder and CEO of enterprise software company Solo.io, told TechCrunch she sees customers switching from proprietary models to open-source models, AI systems whose underlying code is publicly available, installed on servers the company physically controls. Open models now handle 29% of traffic through Vercel's AI routing tools. We reported in July that open-source AI adoption has already reached half the Fortune 500, in a story that found the companies leading that shift aren't only the big tech names you'd expect.
The plain judgement from this beat: Nadella's warning is real, but it's also perfectly timed for Microsoft's cloud sales pitch. The data-leakage risk is genuine, the remedy he proposes happens to benefit Azure, and most enterprises won't act until a competitor's product turns up in a rival's hands. That's when the fine print stops being abstract.
Watch for: AI service agreements that include phrases like "we may use your data to improve our services" without a clear opt-out. That clause, buried in the small print, is the mechanism Nadella is warning about.
