According to TechCrunch, Microsoft CEO Satya Nadella has warned that companies using outside AI models are effectively paying for intelligence twice — once for tokens, and again by handing over their most valuable data. As firms run models from labs like OpenAI and Anthropic, Nadella argues those labs gain access to sensitive workflows, prompts, and the corrections employees make when an AI gets something wrong — a goldmine of industry knowledge given away for free. His concern is that model makers could eventually use that knowledge to compete with their own customers. Nadella urged businesses to keep ownership of their data, build private learning environments, and add orchestration layers that make switching between AI providers easy rather than getting locked into one.