Anthropic Secures $380 B Valuation While Accusing Three Chinese Labs of Massive Claude Data Theft
Updated (3 articles)
Funding Boost Propels Anthropic to $380 B Valuation On 24 February 2026 the U.S. startup announced a capital raise that lifted its post‑money valuation to $380 billion (€322 billion) [1]. Founder and CEO Dario Amodei, a former OpenAI researcher, led the round, positioning Anthropic among the world’s most valuable AI firms. The funding comes as the company intensifies its public stance on AI safety and geopolitical competition [1].
Anthropic Alleges Large‑Scale Claude Distillation by DeepSeek, Moonshot, MiniMax In a press release the same day, Anthropic claimed that Chinese labs DeepSeek, Moonshot and MiniMax (spelled “Minimax” by CNN) executed systematic illicit distillation of its Claude model [1][2]. The allegation includes the creation of over 24,000 fraudulent accounts and the harvesting of more than 16 million Claude interactions to train rival models [2]. Anthropic stresses that such distillation breaches its licensing terms and that Claude is not officially offered in China [2].
U.S. Policy Response Emphasizes Export Controls and National Security Amodei’s late‑January paper warned of advanced‑AI risks, framing the current dispute as a national‑security issue [1]. Anthropic urges tighter U.S. export restrictions, arguing that the stolen capabilities combined with advanced chips could erode America’s AI lead [2]. The accusations surface weeks before former President Donald Trump’s scheduled state visit to China, heightening diplomatic pressure on Washington to act [1].
Market and Pentagon Reactions Highlight Growing AI Rivalry Following the announcements, IBM’s shares plunged more than 13 % in a single session, reflecting investor anxiety about competition from newer AI entrants [1]. The Pentagon has opened an ethical review of Anthropic’s involvement in a U.S. operation that removed Venezuelan President Nicolás Maduro, underscoring defense‑sector concerns [1]. CNN’s report does not mention these market or defense impacts, focusing solely on the technical theft allegations [2].
Sources
-
1.
Le Monde:Anthropic at the Center of the U.S.–China AI Conflict:**Provides valuation details, outlines accusations against three Chinese labs, reports IBM’s stock drop, notes Pentagon ethical review, and links the timing to Trump’s China visit, emphasizing geopolitical stakes.
-
2.
CNN:Anthropic Accuses Chinese AI Labs of Illegal Claude Distillation:**Concentrates on the scale of data theft (24 k fake accounts, 16 M interactions), safety‑guard concerns, parallels OpenAI’s similar claims, and calls for stricter export controls, offering a technical and policy‑focused perspective.
Related Tickers
Timeline
Jan 2025 – DeepSeek releases R‑1, the first Chinese open‑source large language model, sparking rapid adoption by global firms and setting the stage for a wave of Chinese AI exports. [1]
Dec 2025 – A Stanford study concludes that Chinese open‑source models have caught up with or surpassed global peers in capability and user adoption, attributing the surge to strong government backing. [1]
Jan 2026 – Pinterest integrates Chinese open‑source models—including DeepSeek, Alibaba’s Qwen and Moonshot’s Kimi—into its recommendation engine, branding the move a “DeepSeek moment” and turning the platform into an AI‑powered shopping assistant. [1]
Jan 2026 – Airbnb adopts Alibaba’s Qwen for its AI customer‑service agent, praising the model as “very good, fast and cheap,” while other U.S. firms cite similar cost and accuracy advantages over proprietary alternatives. [1]
Jan 2026 – Anthropic CEO Dario Amodei publishes a late‑January paper warning of existential risks from advanced AI, framing the debate that soon draws his company into a U.S.–China technology clash. [3]
Feb 2026 – Anthropic publicly accuses DeepSeek, Minimax and Moonshot AI of illegally distilling Claude’s capabilities, alleging the creation of over 24,000 fake accounts and harvesting 16 million Claude interactions to train their models. [2][3]
Feb 2026 – Anthropic’s new funding round lifts its post‑money valuation to $380 billion, placing it among the world’s most valuable AI firms and intensifying scrutiny of its role in the emerging AI rivalry. [3]
Feb 2026 – IBM’s shares tumble more than 13 % in a single session as investors react to the unfolding AI competition and fears that Chinese models could erode the market share of legacy U.S. tech giants. [3]
Feb 2026 – OpenAI submits a memo to the U.S. House Select Committee on China alleging that Chinese labs illegally distilled its ChatGPT models, echoing Anthropic’s claims and prompting calls for tighter export controls. [2]
Feb 2026 – Former UK deputy prime minister Sir Nick Clegg tells the BBC that the United States fixates on superintelligence while China “democratizes” AI, highlighting the broader strategic tension framing the AI race as democracy versus autocracy. [1]
Mar 2026 (planned) – Former President Donald Trump schedules a state visit to China, a trip that will occur weeks after Anthropic’s theft accusations and is expected to amplify diplomatic friction over technology and AI security. [3]