Strengthening our defenses against distillation attacks

Key points

• In recent months, we’ve made substantial progress on reducing the effectiveness of distillation attacks on Claude using new methods, as we outline in this blog.1
• Distillation attacks are covert, industrial-scale campaigns to extract the capabilities of one model and replicate them in another, without the authorization of the original model’s developers.1
• First, distillation enables companies to build models that approach the capability frontier, but which have, in practice, been released without proportionally robust safeguards.1
• The US Center for Advancing Innovation and Standards for Super Intelligence (CAISSI) has since assessed GLM-5.3 to be the most cyber-capable open-weights model released to date.1

Sources (1)

• [1]Strengthening our defenses against distillation attacks
  Anthropic Research · Oct 12, 12:00 AM

  Strengthening our defenses against distillation attacks

  In recent months, we’ve made substantial progress on reducing the effectiveness of distillation attacks on Claude using new methods, as we outline in this blog.

Extractive summary: sentences quoted from the sources.