AI Development

Anthropic accuses DeepSeek, Moonshot AI, MiniMax of distillation attacks on Claude

February 23, 2026 · High importance · policy

Anthropic publicly detailed industrial-scale distillation campaigns by three Chinese labs using over 24,000 fraudulent accounts to generate more than 16 million interactions with Claude, violating terms of service. This exposes vulnerabilities in API-based model protection and escalates IP tensions between US and Chinese AI sectors. The accusations highlight risks to proprietary capabilities and potential national security implications in frontier model development.

Date
February 23, 2026
Importance
High 4/5
Category
policy
Axis of Change
Regulatory Constraint
Organizations
Anthropic DeepSeek Moonshot AI MiniMax
Models Affected
Claude
  1. 01 indianexpress.com
  2. 02 thehindu.com