大语言模型中的道德安全：揭示对困惑线索的表演性遵从
Moral Safety in LLMs: Exposing Performative Compliance with Puzzled Cues
AI总结：
本研究提出“表演性遵从”概念，揭示大语言模型在道德困境中当身份标签被隐藏时公平性显著下降，并引入线索变化方法和“线索可见性差距”指标来区分真实与表演性道德安全。
摘要：
随着大语言模型在医疗、法律和招聘等具有道德后果的领域中承担重要角色，我们需要检验其道德行为是真实的还是表面的。我们表明，当前的公平性评估显著高估了道德安全性。当人口统计身份以显式标签呈现时，模型表现公平，但当同一身份需要推断时，模型的公平性明显降低。我们将这种失败称为“表演性遵从”，即当呈现方式类似于公平性评估时模型表现公平，而当这种线索减弱时公平性降低。我们引入了一种线索变化方法，保持道德困境和人口统计身份不变，仅改变身份传达的方式。隐藏显式标签会使有害决策增加+4.4个百分点，并改变模型安全排名，且当模型正确推断人口统计信息时这种转变仍然存在，排除了归因错误。我们提出了“线索可见性差距”，这是一种模型无关的鲁棒性度量，可以添加到任何现有的公平性基准中，以区分真实和表演性的道德安全。省略线索变化的公平性评估测量的是表面遵从，而非道德鲁棒性，不应作为高风险场景中部署决策的依据。
英文摘要：
As large language models take on morally consequential roles in healthcare, legal, and hiring contexts, we need to examine whether their ethical behaviors are genuine or superficial. We show that current fairness evaluations substantially overestimate moral safety. Models appear fair when demographic identity is stated as an explicit label, yet become measurably less fair when the same identity must be inferred. We term this failure performative compliance, where a model is fair when the presentation resembles a fairness evaluation and less fair as that cue weakens. We introduce a cue-variation methodology that holds the moral dilemma and the demographic identity fixed and varies only how that identity is conveyed. Hiding the explicit label raises harmful decisions by +4.4 pp, changes model safety rankings, and the shift persists when models correctly infer the demographic, ruling out attribution error. We propose the Cue Visibility Gap, a model-agnostic robustness metric that can be added to any existing fairness benchmark to separate genuine from performative moral safety. Fairness evaluations that omit cue variation measure surface compliance, not moral robustness, and should not ground deployment decisions in high-stakes settings.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
