NEWS · COMPANIES · #2276
Anthropic says new defenses cut distillation data from Claude by ~90–99%
Anthropic describes a set of expanded defenses to hinder distillation attacks on its Claude models, reporting early evidence that the measures have reduced the amount of data harvested by distillers by roughly 90–99%. The company says it has blocked active extraction tactics, sacrificed revenue and trillions of tokens of paying traffic, and changed model outputs (e.g., summarizing reasoning) to make transcripts less useful for training stolen models, while noting trade-offs for customers and integrations.
KEY POINTS
- Anthropic describes a set of expanded defenses to hinder distillation attacks on its Claude models, reporting early evidence that the measures have reduced the amount of data harvested by distillers by roughly 90–99%.
- The company says it has blocked active extraction tactics, sacrificed revenue and trillions of tokens of paying traffic, and changed model outputs (e.g., summarizing reasoning) to make transcripts less useful for training stolen models, while noting trade-offs for customers and integrations.
- Distillation enables illicit replication of frontier model capabilities and can produce less-safe open models that pose misuse and national-security risks.
WHY IT MATTERS
Distillation enables illicit replication of frontier model capabilities and can produce less-safe open models that pose misuse and national-security risks.