Three of the most valuable AI companies in the world are trying to do something they've never done before: coordinate against a shared threat. According to a Bloomberg report published April 6, OpenAI, Anthropic, and Google are forming a coalition to address the alleged copying of their proprietary AI models by Chinese competitors.
The technical mechanism at issue is distillation - a process where a smaller "student" model is trained on the outputs of a larger "teacher" model, absorbing its capabilities without access to the original training data. Legitimate research uses this technique openly. The coalition's concern is that Chinese companies have been doing this with commercial models without permission, essentially using billions of dollars of American R&D to shortcut their own development cycles.
This isn't a new accusation. Earlier in 2026, OpenAI alleged that DeepSeek's R1 model showed evidence of distillation from GPT-4. What's new is that OpenAI, Anthropic, and Google are choosing to respond together rather than separately - notable given how fiercely these companies compete for customers, talent, and compute.
Bloomberg's report doesn't detail exactly how the coalition will operate. Possible approaches include shared detection methods for identifying whether a model was trained on another's outputs, joint lobbying for tighter US export controls on AI technologies, or coordinated legal strategies.
The practical difficulty here is real. Detection requires forensic analysis that isn't always conclusive - a model trained on distilled outputs can be hard to distinguish from one trained independently. Open-weight models, which anyone can download freely, can be distilled by anyone, anywhere, with no paper trail. The coalition signals serious intent, but enforcement against foreign actors operating outside US legal jurisdiction has no clean answer.