Kai-Fu Lee says Anthropic owes him $3,000 amid its complaints about Chinese IP infringement

The claim

It has been reported that Kai-Fu Lee (李开复) publicly criticized Anthropic for “complaining about Chinese companies infringing rights” while allegedly still owing him $3,000 in manuscript fees. The comment, surfaced by Phoenix Technology (凤凰科技), did not include details of the work involved or the timing of the unpaid fee. Anthropic has not publicly addressed the claim, according to the report. A minor sum? Perhaps. But the juxtaposition is striking.

Who is Lee—and why it matters

Lee is one of China’s most prominent AI figures: a former head of Google China, founder of venture firm Sinovation Ventures (创新工场), and founder of model developer 01.AI (零一万物), which is building domestic large language models to compete with Western offerings. His remarks carry weight in China’s tech community and increasingly shape the narrative around cross-border AI competition. When he calls out a leading U.S. lab, the industry listens.

The other side: Anthropic’s posture

Anthropic, a U.S. AI company backed by Amazon and Google and known for its “AI safety” framing, has taken public stances on responsible AI development and has reportedly criticized misuse and intellectual property violations by developers, including in China. That stance sits uncomfortably next to Lee’s unpaid-fee allegation, if true. Is this a mere online spat—or a symptom of deeper frictions over IP norms, attribution, and compensation in the AI era?

Bigger picture

The dust-up lands amid intensifying U.S.–China tech rivalry: export controls on advanced chips, tightening compliance expectations, and a global scramble to commercialize foundation models. Cross-border grievances—whether about model training data, licensing, or honoraria—are increasingly politicized. The $3,000 figure may be symbolic, but the message is not: trust is thin, and even small disputes can amplify broader anxieties about fairness and reciprocity in AI.