Alibaba ran 29 million illicit Claude exchanges, Anthropic tells Senate
Anthropic named Alibaba's Qwen lab as the source of 28.8 million unauthorized Claude exchanges via 25,000 fake accounts between April and June — the largest distillation campaign attributed to a single Chinese firm, targeting software engineering and agentic reasoning. A bipartisan Senate amendment would blacklist firms that harvest US AI outputs; with Anthropic filing confidentially for an IPO at a $965B valuation, enforceability of that border is now a public-market risk as well as a policy one.
Source: thenextweb.com ↗
operators linked to Alibaba's Qwen AI lab used nearly 25,000 fraudulent accounts to extract Claude's capabilities between April and June
Anthropic letter to US senators
Why this matters
- → 29M unauthorized exchanges expose massive IP theft targeting US AI leadership
- → Senate amendment would penalize distillation, affecting China's model-building economics
- → Anthropic's IPO timing makes enforcement a public-market risk, not just policy
China's AI theft escalates