精选理由
Anthropic 的 Opus 5 在防提示注入上做到了史上最强，红队测试都很难攻破，安全团队值得关注。
Anthropic 的 Claude Opus 5 在提示注入（prompt injection）防护上取得显著进展。根据官方系统卡和红队评估，Opus 5 是 Anthropic 迄今最难以被成功提示注入的模型。这一结果基于多项 PI 评估和对抗测试，标志着大模型安全性的一次重要提升。
原文 · Simon Willison’s Weblog
Quoting Boris Cherny
More than any of these eval scores, what is most exciting to me is something else: Opus 5 is our least prompt injectable model yet. It is a bit buried in the system card, but across PI evals and red teaming, Opus 5 is very hard to prompt inject successfully. — Boris Cherny , here's that System Card section , page 73 Tags: prompt-injection , anthropic , claude , generative-ai , ai , llms , boris-cherny