Anthropic says Alibaba must be punished for largest Claude cloning attack

Anthropic has accused Alibaba of running the largest-ever attempt to clone its Claude AI, using nearly 25,000 fake accounts to conduct over 28.8 million exchanges between April and June 2026. The goal was to steal Claude's capabilities without paying the massive R&D costs required to build a competing model. Anthropic is now urging Congress to update antitrust laws, tighten chip export controls, and penalize Chinese labs engaging in these distillation attacks.

(A distillation attack is when someone uses an existing LLM’s outputs to train a new model that imitates or extracts the behavior, capabilities, or proprietary knowledge of the original model without authorization.)