AI news · Monday, July 13, 2026

Microsoft CEO warns businesses are paying for AI with proprietary data

The relationship between Big Tech and AI labs is turning sour, with Microsoft CEO Satya Nadella publicly warning businesses that using proprietary models like Claude or GPT is a double-edged sword. He argues companies are essentially paying twice: once in cash for token usage, and again by handing over their sensitive business data, which labs then use to improve their models. This process—often called 'exhaust' learning—creates an intellectual property drain that Nadella calls hypocritical, especially given that labs lean on fair-use arguments to scrape the open web for their own training data. He is nudging enterprises toward 'proprietary learning environments' on the cloud where they can retain ownership, though this also happens to align with Microsoft’s own Azure infrastructure goals. This shift is already manifesting as companies increasingly run open-source models 'on-prem' to keep their data under lock and key, a move supported by specialized software firms like Solo.io.

OpenAI is simultaneously facing a high-profile legal battle with Apple, which accuses the startup of systematically poaching employees and stealing trade secrets. Apple’s 41-page complaint details allegations that OpenAI staff were instructed to bring Apple hardware prototypes to job interviews and that former Apple engineers exploited a 'zero-day' vulnerability—a previously unknown security bug—to siphon files from Apple’s servers after they had already jumped ship. The suit paints a picture of a culture that prioritized getting ahead at any cost, marked by internal texts where departing engineers jokingly celebrated their ability to bypass Apple’s network security. OpenAI has denied these claims, asserting it has no interest in third-party trade secrets.

Amidst the corporate drama, the reality of AI-driven job displacement remains a point of intense friction. Over 200 leaders, including former Google CEO Eric Schmidt and Nobel laureates, signed a letter warning that AI’s transformation could rival the Industrial Revolution, threatening large-scale job losses unless policymakers act now. This anxiety is hitting the workforce directly, with software engineers pivoting to entirely different industries like nursing after months of layoffs and 'AI overload' in job interviews. Even internal management is struggling to balance the tech; employees are reporting that their managers are forcing them to credit AI chatbots for months of human labor, often stalling raises or promotions because executives assume the machine did the work for free. Meanwhile, some in the industry are taking a more cynical view, with Palantir CEO Alex Karp noting that the real danger isn't necessarily mass joblessness, but a 'complete decoupling' of wealth where a handful of AI leaders become 100 times richer than they already are while the average person's gains remain stagnant.

The quick hits

Apple is suing OpenAI for allegedly poaching employees and stealing trade secrets to build a competing hardware device — a messy escalation that highlights the breakdown of their previous partnership.
Microsoft CEO Satya Nadella warned businesses they are essentially 'paying twice' for AI by handing over proprietary data to proprietary models — a push to keep enterprise data on private cloud systems.
Over 200 economists and tech leaders signed a letter urging policymakers to build guardrails against 'large-scale job displacement' as AI development accelerates faster than the economy can adapt.
Employees report that managers are forcing them to credit AI for work they performed themselves, leading to lower raises and stalled promotions because bosses assume the software did the heavy lifting.

Microsoft CEO warns businesses are paying for AI with proprietary data

The quick hits

Sources