Satya Nadella has warned companies that they pay twice when they use AI models from large labs: once with money, and again by handing over sensitive business knowledge.
While Silicon Valley enthusiasts worry that major AI labs act like Trojan horses, Nadella says the danger is more direct. Enterprises feed proprietary data into these systems to make them work better. The model makers then use that information to train their own models or compete with their paying customers.
The warning
Nadella published this view in a blog post on Monday. He argues that users unknowingly give away valuable data while paying for token usage.
“You essentially pay for intelligence twice, once with money, and again with something even more valuable: the proprietary knowledge you must reveal to make that intelligence useful. The better you want the model to perform, the more of that knowledge you have to feed it!”
He states that companies are teaching models the specific nuances of their own businesses. Every prompt a person writes, every tool an agent uses, and especially every correction made when the model errs, becomes institutional know-how.
“This is the kind of knowledge a competitor could never buy,” Nadella writes. “Yet enterprises are handing it over.”
Distillation and hypocrisy
Nadella argues that if AI companies can freely scrape the internet to train their models, it is fair that enterprises should be allowed to study those models in return. This process is known as distillation. It involves using a model’s outputs to understand how it works and to train a new, often cheaper, model based on those insights.
In February, Anthropic accused Chinese open source models of sending millions of prompts to Claude to improve their own systems. They urged the US government to crack down on such export controls.
Nadella’s point is that model makers cannot have it both ways. It is hypocritical for them to freely train on public data while restricting others from doing the same to their own models.
“While the great innovation that comes from model providers having fair use rights to train models on public data is needed, I find it ironic that the status quo is to then turn around and impose restrictive terms on distillation,” Nadella writes.
He is particularly concerned when model makers reserve the right to learn from customer usage and interaction data.
A solution
Nadella wants companies to retain ownership of their data, including prompts and feedback. He suggests building proprietary learning environments on the cloud, where data is likely already stored.
He also wants companies to build orchestration layers. These allow businesses to switch between AI models from different providers rather than being locked into one. Tools like AI gateways, which let companies do exactly this, have become increasingly popular.
While Nadella does not use the word open-source as the method for retaining ownership, this is an obvious subtext. There is another subtext as well.
Large companies, many of which still have their own data centres in addition to using the cloud, are already moving to open source models installed on their own premises. Idit Levine, founder and CEO of Solo.io, says she is seeing this shift with her own customers.
After experimenting with proprietary model makers, they start asking themselves: “Can I take an open-source model and run it on-prem? It will do almost 90% of what the big one’s doing. It will cost way less,” she tells TechCrunch. “They understand that, and they can control it.”
Solo.io’s technology was selected last year as the tech powering the Linux Foundation’s Agentgateway project. Her company counts enterprises like T-Mobile, ADP and SAP as customers. She sees companies increasingly installing on-premise open source models and sees it as the next big wave in enterprise AI use.
She is not alone. Vercel, best known as a platform for building and hosting websites, and OpenRouter, a company that helps developers route requests across different AI models, are both seeing a surge in traffic to open-source models. In fact, open models accounted for 29% of all traffic routed through Vercel’s gateway last month.
With the CEO of Microsoft, a company that has invested in both OpenAI and Anthropic, now openly urging enterprises to be wary of using proprietary models, we will bet this trend continues to grow.
“In consuming intelligence, you are creating intelligence. And what you create should belong to you,” Nadella writes.
