"Anthropic accuses DeepSeek, MiniMax, & Moonshot of misusing Claude AI model. Investigations revealed 24,000 fake accounts & over 16M interactions. Concern
**Anthropic Accuses Three Chinese AI Companies of Misusing Claude Model**
In a significant development, U.S.-based AI research company Anthropic has accused three Chinese AI companies - DeepSeek, MiniMax, and Moonshot - of misusing its Claude AI model. The allegations were made in an announcement on Monday, following investigations that uncovered "industrial-scale campaigns" involving the creation of approximately 24,000 fraudulent accounts and over 16 million exchanges with Claude.
**Alleged Misuse of Claude Model**
- Approximately 24,000 fraudulent accounts created
- Over 16 million exchanges with Anthropic's Claude AI model
- Accusations of "distilling" Claude, a training method also used for illicit purposes
- Concerns over the transfer of capabilities to military, intelligence, and surveillance systems
- Unprotected capabilities potentially enabling offensive cyber operations, disinformation campaigns, and mass surveillance
**Targeted Company: DeepSeek**
- Over 150,000 exchanges with Claude
- Targeted reasoning capabilities
- Accusations of generating censorship-safe alternatives to politically sensitive questions
- Previously accused by OpenAI of "ongoing efforts to free-ride on the capabilities developed by U.S. frontier labs"
**Other Companies Involved: Moonshot and MiniMax**
- Moonshot had over 3.4 million exchanges with Claude
- MiniMax had over 13 million exchanges with Claude
**Call for Industry Action and Regulation**
Anthropic is urging other members of the AI industry, cloud providers, and lawmakers to address the issue of distillation. Suggestions include implementing restricted chip access, which could limit model training and the scale of illicit distillation.
Source: Read Original Article
Post a Comment