Bloomberg reported on 24 June that Anthropic accused Alibaba of waging a large-scale effort to “illicitly” access its artificial intelligence systems using thousands of fraudulent accounts, according to a letter the US developer sent to several senators and White House officials [E1].
Anthropic described the operation as the “biggest attempt so far by a Chinese company to piggyback on the work of top US labs,” Bloomberg wrote, saying operators linked to Alibaba’s AI lab targeted prized capabilities in software engineering and agentic reasoning [E1][E2]. The letter, dated around 10 June and cited in the reporting, alleged 25,000 fraudulent accounts conducted 28.8 million exchanges with Anthropic’s systems between 22 April and 5 June 2026 [E2]. Those figures rest on Anthropic’s own forensic account; neither a court nor an independent auditor has verified them.
Alibaba has not been charged, and Bloomberg’s account rests on Anthropic’s allegations rather than an adjudicated finding. The dispute lands after the New York Times reported that Anthropic shut down its two most powerful systems following an “unexpected demand from the U.S. government to cut access to it” in mid-June [E3]. Anthropic’s letter also referenced earlier, smaller campaigns it attributed to other Chinese developers, framing the Alibaba episode as an escalation in a distillation contest that US labs have been flagging for months [E1].
Chinese startups continue to release capable offerings at lower price points even as US developers tighten access. The New York Times reported on 25 June that Z.ai shipped a new system nearly as powerful as Anthropic’s top tier but “costs much less to use,” landing quickly on a closely watched popularity leaderboard where “six of the models now on the A.I. leaderboard were developed in China” [E4]. A Silicon Valley founder told the Times that after Washington restricted Anthropic’s strongest tier, “the gap between the U.S. and China is very slim” [E5].
Cost pressure on Western labs is compounding the geopolitical friction. CNBC reported on 26 June that enterprises are abandoning unconstrained token spending in favor of routing and cheaper providers, with one startup chief executive switching all traffic off Anthropic and telling the outlet savings were “a matter of survival for the business” [E6]. Analyst Gil Luria told CNBC that “current growth rates for Anthropic and OpenAI are the fastest they will ever be,” while warning that “some of their largest enterprise customers may start limiting their out-of-control token spend” [E7].
Both Anthropic and OpenAI filed confidential initial public offering paperwork in early June, CNBC noted, as finance teams confront bills that many chief financial officers did not plan for in annual budgets [E6]. Anthropic last disclosed a $47 billion annualized run rate in May, CNBC reported, even as consultants described a broad spend crunch, with one executive saying “Everybody is experiencing the same spend crunch on AI” among clients delaying large commitments until return on investment is clearer [E8].
Microsoft, Amazon, and Google are pitching lower-cost in-house or routed alternatives, CNBC added, a competitive flank that pressures pure-play labs to defend pricing as Chinese open-weight offerings gain US adopters despite security and provenance concerns [E6][E4].
For now the fight is two-fronted: Anthropic has put Alibaba on the record in a Senate-facing letter, but proof of state-sponsored theft remains allegation, not verdict [E1][E2]. Meanwhile cheaper Chinese releases and enterprise routing are reshaping who pays frontier prices at all, a squeeze that may outlast any single distillation campaign [E4][E7].