Google fights order to share search data with OpenAI and rivals in US antitrust case

This appeal is about more than Google’s ad business; it goes to the heart of how much access rival AI firms can have to the data exhaust of the world’s dominant search engine. Judge Mehta’s remedy, as described, would force Google to share certain search data with competitors like OpenAI, effectively treating parts of its clickstream as regulated infrastructure rather than proprietary fuel for models and ranking algorithms. Google is pushing back hard, arguing that such sharing would expose trade secrets and downplaying the degree to which defaults rather than user choice drive its dominance. ([timesofindia.indiatimes.com](https://timesofindia.indiatimes.com/technology/tech-news/google-to-court-cannot-share-data-with-chatgpt-maker-openai-and-others-the-ruling-ignores-the-basic-reality-that-/articleshow/126636269.cms))

For the AGI race, the stakes are obvious. Search logs and ad-click histories encode an unparalleled map of human intent and preference—exactly the kind of signal labs want to train better agents and recommendation systems. If courts ultimately force more open access to that data, it could dramatically narrow the moat between Google and labs like OpenAI, Anthropic or xAI. Conversely, if Google wins, it reinforces a world where a few vertically integrated giants can combine their own data monopolies, models and distribution channels to outpace pure-play AI labs.

The case also previews a regulatory future where AI remedies go beyond model-level rules into structural constraints on who owns and controls the input data that makes AGI training possible in the first place.