tech
Anthropic is accusing China's Alibaba of exploiting its AI models in a large-scale attack
Anthropic said the Chinese company interacted with Claude millions of times through 25,000 fraudulent accounts in recent months.
TL;DR
- Anthropic accuses Alibaba of a large-scale "distillation attack" on its AI models.
- The attack involved 28.8 million exchanges with Claude through nearly 25,000 fraudulent accounts.
- The goal was to illicitly extract Claude's capabilities to train Alibaba's own AI models.
- Anthropic is urging lawmakers to implement legislation to prevent further attacks and limit China's access to advanced US computing infrastructure.
- This accusation follows recent US government actions against Alibaba, including export controls on AI models and a Pentagon blacklist.
- Alibaba has denied previous allegations and recently sued the US government over its designation.