Satya Nadella warns companies they are giving AI labs their most valuable secrets
Microsoft CEO Satya Nadella has issued a surprising warning to businesses, cautioning that using proprietary AI models forces them to pay twice. He argues that companies pay once with money and a second time by exposing valuable, proprietary institutional knowledge through prompts, usage patterns, and error corrections. According to Nadella, AI model makers like OpenAI and Anthropic are acting like Trojan horses by absorbing this business "exhaust" to improve their own systems. He calls it ironic that these labs rely on fair use to scrape public data but restrict customers from distilling their models. To combat this, Nadella urges enterprises to build their own orchestration layers and transition to open-source models hosted on secure cloud infrastructures.