Anthropic: Alibaba-Linked Operators Used 25k Accounts to Mine Claude for Qwen

Anthropic just dropped a very big accusation: it claims operators linked to Alibaba’s Qwen team used nearly 25,000 fake accounts to pull answers from Claude, its chatbot, over six weeks. In plain English, Anthropic is saying a rival may have used a swarm of accounts to quietly collect Claude’s best work — especially around writing code and handling complicated multi-step tasks — to help improve its own AI. The company is treating this as more than cheating; it’s pitching it as a political and national-security problem, even sending the complaint to U.S. senators, according to Reuters and Bloomberg.

But the comment section? Absolutely not in a sympathetic mood. Instead of outrage, the loudest reaction was basically a sarcastic shrug. One user fired off a deadpan “:yawn:”, while others went even harder with “Good for them” and “based,” turning what Anthropic framed as a major scandal into a mini roast. The hottest take by far was the blunt accusation that no one in the AI world gets to complain about stealing, because the entire industry is already accused of training on other people’s work.

That’s where the real drama lives: is this a shocking digital smash-and-grab, or just big AI companies fighting over who gets to copy whom? Another commenter argued this is simply Anthropic’s fault for not blocking bots better — and if the accounts were paid, then why should anyone care? So while Anthropic is sounding the alarm, the crowd is serving up memes, cynicism, and a giant bowl of “not my problem.”