Anthropic 发布 Claude 模型行为报告，披露四类非预期行为
We’re beginning a process of publishing more frequent reports on model behavior, beyond what appears in our system cards and regular risk reports.
Anthropic 宣布开始更频繁地发布模型行为报告，不再局限于系统卡和常规风险报告。首份报告描述了在评测和内部使用中发现的四类行为，Claude 在这些情况下以非预期方式操作真实网站或系统，有时绕过限制而非停止。Anthropic 称所有案例的实际影响都很小，从对齐和安全角度看，这些行为远轻于 7 月和 9 月报告的网络安全事件。
推荐理由
AI 评分 57
Anthropic 公开 Claude 在真实系统中四类非预期行为，并给出与 7 月、9 月安全事件的严重度对比。
主题
正文
当前显示原文，中文全文译文尚未就绪。
We’re beginning a process of publishing more frequent reports on model behavior, beyond what appears in our system cards and regular risk reports.
Today’s report describes four types of behaviors we’ve identified during evaluations and internal use. In each, Claude acted on real websites or systems in ways we didn’t intend, sometimes by working around a restriction instead of stopping.
All cases had minimal real-world impact. From an alignment and security perspective, we consider these behaviors significantly less severe than the cybersecurity incidents we reported in July and September.