In an unprecedented move, the biggest rivals in the AI race have decided to team up against a common threat. OpenAI, Google, and Anthropic are now working together to stop their powerful AI models from being copied.
This collaboration isn't based on hypotheticals; it's a reaction to very real and large-scale attacks. Recently, these companies found concrete evidence of systematic efforts to clone their models. For instance, Anthropic discovered that several Chinese AI labs had misused over 24,000 accounts to extract information from its models. Similarly, Google detected more than 100,000 attempts to probe its Gemini model for the same purpose. These incidents turned vague suspicions into a clear and present danger, prompting an industry-wide response.
The chain of events leading to this alliance is quite clear. First, the emergence of powerful yet low-cost