[KongchangAI]
Product网络验证计划 / Cyber Verification Program

CVP

A controlled AI red-team authorization mechanism established by Anthropic. It permits vetted and approved organizations to conduct cybersecurity probing and red-team testing on AI models within clearly defined scopes, serving as a key compliance channel for safety governance by large model providers.

Timeline (last 90 days)

Nov 14

Some users say that CVP personal applications now only require KYC and no longer require KY or its certification.

Unverified50%

All Facts (1)

Source Articles