Anthropic News · 29 天前精选模型评分63
Anthropic 扩展 Cyber Verification Program，分三档向安全团队开放高级模型能力
Expanding the Cyber Verification Program
导读
Anthropic 推出扩展版 Cyber Verification Program（CVP），把此前的 Project Glasswing 和 CVP 合并，按防御、红队、专项三档向通过审核的安全团队开放 Claude Opus 5.5、Claude Sonnet 5.5、Claude Mythos 5.1 等模型并降低安全拦截。
推荐理由
原文给出三档访问权限的划分和基准测试结果，读者可据此判断安全团队能获得多少模型能力。
分类置信度
高
- 官方或准官方来源，且无冲突类别。
- 正文、可信日期、主分类和证据摘录完整。
来源级别：官方一手
应用领域：政务与公共事务
事件类型：产品更新
证据摘录
- “We’re launching a new, expanded version of our Cyber Verification Program (CVP), which makes advanced cyber capabilities and reduced blocking classifiers available to qualifying security professionals.”
- “We aim to respond to applications within a few days.”
- “Given the increased eligibility requirements and security controls, we expect applications in this tier to take a few weeks to review.”
- “Currently, this tier is for organizations only; individual researchers are not eligible.”
- “CVP is only available on Amazon Bedrock for customers eligible for Enterprise Frontier Safeguards .”
来源：Anthropic News · anthropic.com