Member-only story
Anthropic Caught Three Chinese Labs Stealing 16 Million AI Conversations. The Internet Took China’s Side.
24,000 fake accounts, a $1.5 billion copyright settlement, and a hypocrisy debate that’s drowning out the actual national security crisis.
Ahmed M. Abdelfattah12 min read·Just now--
24,000.
That’s how many fake accounts three Chinese AI labs were operating simultaneously on Anthropic’s platform when the company’s threat intelligence team mapped the full scope of what was happening.
Not a few researchers testing boundaries. Not a startup cutting corners. Twenty-four thousand coordinated accounts, generating over 16 million exchanges, running through proxy networks designed to make industrial-scale extraction look like normal customer traffic.
On Monday, Anthropic named names: DeepSeek, Moonshot AI, and MiniMax. All based in China, where Anthropic’s services are explicitly banned for legal, regulatory, and security reasons. All allegedly using a technique called distillation to capture Claude’s capabilities at a fraction of the time and cost it took to develop them.
Jacob Klein, Anthropic’s head of threat intelligence, was blunt: “We have high confidence these labs were conducting distillation attacks at scale.”
