Anthropic Strengthens Defenses Against Distillation Attacks
Anthropic says it is adding stronger protections against distillation attacks, where someone tries to copy an AI model's abilities. The move aims to guard the company's models from this kind of misuse.
Quick glossaryAnthropic, distillation attacks
- Anthropic
- A US AI company best known for its Claude chatbot, founded by former OpenAI researchers.
- distillation attacks
- Attempts to copy a stronger AI's skills by repeatedly asking it questions and learning from its answers.
Burning Potato score: 33%