Anthropic accuses Alibaba of 'illicitly' accessing its Claude model

By Financial Times · June 24, 2026.

The Financial Times reports, in its daily FirstFT newsletter, that Anthropic has accused Alibaba of having 'illicitly' accessed Claude, the US company's artificial intelligence model. The article is restricted to FT subscribers and it has not been possible to obtain the full text, so this summary is limited strictly to what appears in the headline's metadata.

The term 'illicitly' used in the FT's own headline suggests that Anthropic would not be describing an ordinary contractual breach —for example, a violation of terms of use— but rather conduct that, in its view, crosses a more serious legal or ethical line. With no further verified information available, any additional detail about the mechanism, scope or legal consequences of that access would be speculative.

As sector context, tensions between Western AI companies and Chinese competitors over the unauthorized use of models and training data have been mounting throughout 2025 and 2026. OpenAI, for example, has publicly denounced attempts at mass extraction of its models, and several labs have tightened their API access controls and terms of service to limit uses that could benefit competitors. However, applying those precedents to the specific Anthropic-Alibaba case without the article's concrete details would be to venture conclusions unsupported by the available material.

Anthropic is one of the most prominent frontier AI companies of the moment, founded in 2021 by former OpenAI employees and backed by multibillion-dollar investments from Amazon and Google, among others. Its Claude model —currently in several versions, including Claude 3 and later iterations— competes directly with OpenAI's GPT-4o and Google's Gemini in the large-scale language model market. Alibaba, for its part, has developed its own Qwen family of models and operates Tongyi Qianwen as a generative AI platform, which makes any unauthorized access to Claude a matter with potential intellectual property and technological competition implications.

Given the absence of the article's full content, it is not possible to confirm whether Anthropic has filed any kind of legal action, whether the incident has been reported to regulatory authorities, or what Alibaba's official response is. Readers are advised to access the Financial Times article directly to obtain the complete and verified information.