Science & Tech · World
AI Firm Anthropic Reports Success in Cutting Model Distillation Attacks
AI company Anthropic announced that new security measures have significantly reduced "distillation attacks" against its Claude models, reporting a reduction of up to 99%. Distillation attacks are a method used to extract information from AI models.
How outlets across the spectrum covered it
L 0 C 3 R 0
- Centre (3): Business Standard, Reuters, The Hindu BusinessLine