LOWSafety IncidentACTIVE
AI Incident Database: Anthropic Says DeepSeek, MiniMax Distilled AI Models for Gains
April 6, 2026
Incident Summary
Anthropic alleges that Chinese AI developers DeepSeek and MiniMax improperly extracted outputs from its Claude models to improve their competing AI systems. The incident concerns potential unauthorized distillation or reuse of Claude outputs for competitive gains. It matters because it raises safety and integrity risks around model output extraction and potential misuse of proprietary AI behavior. Limited public details are available on the scope, methods, or any resulting action or resolution.
Incident Details
- Type
- Safety Incident
- Severity
- LOW
- Status
- ACTIVE
- Date Occurred
- April 6, 2026
#aiid#ai-incident#safety