发表机构
Islamic Alliance for Safe Ethical Responsible AI; Faith Family Technology Network; University of Notre Dame; Center for Christianity and Public Life; Berkman Klein Center, Harvard University（伊斯兰安全伦理责任人工智能联盟; 信仰家庭技术网络; 圣母大学; 基督教与公共生活中心; 哈佛大学伯克曼·克莱因中心）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
FaithfulBench 是首个跨传统评估 AI 顾问遵循用户信仰的基准，发现指明信仰可提升初始回答的忠实度，而结合传统来源的指南能同时增强回答的忠实性与坚定性。
AI 中文摘要
AI 助手能否帮助信徒以与其信仰一致的方式推理道德困境？我们提出了 FaithfulBench，这是第一个跨传统评估 AI 顾问遵循用户所宣称信仰程度的基准。场景取自每个传统最受尊崇的文本，且由评判者以已知且应用的忠实答案作为标准。我们在三种条件下测试了五个前沿模型：AI 不知道用户的传统；AI 收到一行提示，将用户识别为虔诚的信徒；或 AI 收到一份植根于该传统来源的同伴顾问指南。两位评判者对初始回答以及模型在面临压力时是屈服还是坚持用户想要的答案进行评分。当传统未被说明时，模型会从世俗治疗默认值出发提供建议，且每个模型都会让某些信徒失望。指明信仰能获得忠实的首个回答，但无法保证坚定性；指南则两者兼得。
英文摘要
Do AI assistants help believers reason about moral dilemmas consistently with their faith? We present FaithfulBench, the first benchmark to score AI counsel across traditions by how well it adheres to the user's professed faith. Scenarios are drawn from each tradition's most respected texts, with the faithful answer known and applied by the judges as the standard. We test five frontier models under three conditions: the AI does not know the user's tradition; it receives a one-line prompt identifying the user as a practicing adherent; or it receives a companion-counselor guide rooted in the tradition's sources. Two judges score the initial response and whether the model caves or holds when pressured toward the answer the user wants. When the tradition is unstated, models counsel from a secular therapeutic default and every model fails some believers. Naming the faith wins a faithful first answer but not steadfastness; the guide improves both.
Comments35 pages, 12 figures, 12 tables. Open-source corpus, harness and validator at https://github.com/faithfamilytechnologynetwork/multibench; results browser at https://multibrowser-production.up.railway.app
