Li Kaifu comments on Anthropic’s model distillation accusation: It’s a fuss out of a molehill and I still owe me $3,000 in royalties.

In response to the American AI company Anthropic's recent accusation that Chinese companies "distilled" its AI models, Li Kaifu, CEO of Zero One Thing and Chairman of Sinovation Ventures, issued a public response today. He said that the model distillation behavior was not illegal, and pointed out that Anthropic itself has a history of infringing data collection.

In the public video, Kai-Fu Lee believed that Anthropic’s accusations against model distillation were “making a mountain out of a molehill.” He disclosed the company's record of copyright disputes, pointing out that it had faced a class-action lawsuit from American authors for illegally grabbing millions of books for AI training, and eventually paid a huge settlement of US$1.5 billion (approximately 10.355 billion yuan). As one of the affected writers, he is still waiting for copyright compensation of US$3,000 (approximately 20,709 yuan) from the company.

Previously, Musk criticized Anthropic’s data acquisition methods on social platforms, accusing it of stealing human programmers’ project data for training without authorization. When faced with netizens raising concerns about the data source of his AI product Grok, Musk acknowledged this fact, but emphasized that his team would not show extreme complacency and hypocrisy on the issue of data theft like Anthropic did.