China accused of massive AI distillation theft scheme
Tap title to read story
Anthropic has accused Chinese technology conglomerate Alibaba of organizing what it calls the largest known artificial intelligence distillation attack against its Claude models, according to a June 10, 2026 letter sent to US Senate Banking Committee leaders and shared with the White House. The company alleges operators linked to Alibaba and its Qwen AI lab created about 25,000 fraudulent accounts between April 22 and June 5, 2026, generating roughly 28.8 million interactions with Claude. Anthropic says the activity violated its terms of service and geographic restrictions and was used to train Alibaba’s own AI systems. Alibaba has denied any wrongdoing.
More context
Timeline
- Apr 22, 2026 Fraudulent Claude accounts begin appearing
- Late April 2026 Large-scale interactions with Claude models
- May 2026 Operators expand account creation efforts
- Early Jun 2026 Anthropic detects unusual interaction volumes
- Jun 5, 2026 Fraudulent account creation reportedly ends
- Jun 10, 2026 Anthropic sends letter to US lawmakers
- Mid Jun 2026 Allegations against Alibaba become public
- Mid Jun 2026 Alibaba issues formal denial statement
Why it matters
This alleged AI theft could impact your online safety. If true, Alibaba's actions may have violated privacy norms. It's a good reminder to check the security settings on your AI devices and apps.
JQJO Editorial Desk · Human reviewed · AI assisted. Report a correction.
