Satya Nadella, CEO of Microsoft, published a blog post arguing that companies using proprietary AI models are effectively paying twice, once financially and again by surrendering valuable proprietary business knowledge to model makers through everyday usage. Nadella wrote that AI models learn from user interactions, including corrections and prompts, allowing model providers to absorb institutional knowledge that competitors could never otherwise access.
Nadella argued that it is inconsistent for AI companies to freely train on public internet data while restricting enterprises from studying or “distilling” their own models in return, a practice Anthropic previously accused Chinese AI firms of using against Claude. He said businesses should retain ownership of their data and prompts by building their own proprietary learning environments, potentially on cloud infrastructure, and by adopting orchestration tools that allow easy switching between AI providers rather than relying on a single model maker.
The comments align with a broader shift observed by industry figures, including Solo.io CEO Idit Levine, who said enterprise customers are increasingly moving toward open source models deployed on their own infrastructure as a lower-cost, more controllable alternative to proprietary systems. Platforms including Vercel and OpenRouter have also reported rising traffic toward open source models, suggesting growing enterprise interest in reducing dependence on major AI labs.
