Google has restricted how much Meta can use Gemini, Google's powerful artificial intelligence system. The Financial Times first reported the move in late June 2026, according to Bloomberg.

Meta is one of the biggest companies developing AI in the world. It had been using Gemini as a customer, much like any business might buy software from another company. But Google has now put a cap on that access — though the exact details remain unclear. It could be a slowdown in how much Meta can use it, a limit on which features Meta gets, or something else entirely. Neither Google nor Meta has publicly confirmed the specifics.

Why This Matters

Meta's situation makes this move interesting. The company creates its own AI models and releases many of them publicly so anyone can use them for free. Its Llama 3.1 405B model, released in 2024, competes directly with Google's paid AI services. But at the same time, Meta also buys AI services from companies like Google — a common practice in big organizations that want options.

The pattern is straightforward enough: large technology companies often build their own tools while also paying competitors for theirs. This hedges their bets. But Google's decision to limit Meta suggests that Google views Meta's use of Gemini as potentially problematic.

One possibility is that Meta was using Gemini to help train or improve its own competing models. When one company uses another company's AI system to make its own version better, that's sometimes called model distillation. AI companies have grown concerned about this practice over the past couple of years, and many now have contract terms that forbid it. That could be what happened here.

The challenge is that very little of this is confirmed. The restriction was reported by the Financial Times but neither Google nor Meta has gone on the record with details. We do not know how long this will last, how broad the limits are, or exactly why Google made the decision. Anyone following this story should treat it as a report that needs more evidence before drawing firm conclusions.

The Bigger Picture

The situation between Google and Meta reflects a larger problem in the technology industry right now. Many companies are both cloud customers and direct rivals. Google Cloud sells services to companies that compete with Google's other products. Microsoft does the same thing with its Azure platform. This creates real tension.

Meta makes this tension very visible. By releasing free AI models that anyone can use, Meta is trying to break the industry's reliance on paid services from companies like Google and OpenAI. If Google is now limiting Meta's access to Gemini because of that strategy, it shows the conflict has moved from theory into business reality.

What remains unclear is whether this is a one-time dispute or the start of a larger shift. We may see more AI companies beginning to block or restrict their competitors' access. If so, the question of who gets to buy from which AI company, and under what terms, will become a much more important business and legal issue.