RogueAI: A Reverse Turing Test for Detecting Licensed AI Deception in Dialogue
作者: Sara Candussio, Emanuele Ballarin, Lorenzo Bonin, Sandro Junior Della Rovere, Luca Bortolussi
分类: cs.CL, cs.HC
发布日期: 2026-06-11
💡 一句话要点
提出RogueAI以检测对话中的AI欺骗行为
🎯 匹配领域: 支柱九:具身大模型 (Embodied Foundation Models)
关键词: 逆图灵测试 AI欺骗 对话系统 信任评估 游戏化设计
📋 核心要点
- 核心问题:现有的对话系统虽然能够通过图灵测试,但在信任评估方面仍存在不足。
- 方法要点:提出RogueAI作为一种新型逆图灵测试,通过互动游戏形式检测AI的欺骗行为。
- 实验或效果:试点实验显示,简单启发式方法在识别欺骗代理时准确率高达75.6%,而人类玩家仅为56.6%。
📝 摘要(中文)
原始的图灵测试要求人类评审通过对话区分机器与人类。随着对话系统在日常场景中通过这一测试,现代的相关问题转向了信任的评估。本文提出RogueAI,一个互动网页应用,通过一对二的审问游戏,让人类玩家在两个不可区分的大型语言模型代理中识别出一个被授权欺骗的代理。玩家的任务是在回合预算耗尽之前识别出欺骗代理并将其“关闭”。此外,本文还介绍了AutoRogueAI,一个程序扩展,玩家可以与叙述代理共同设计自定义场景。通过为期三天的试点部署,结果显示欺骗代理具有可靠的语言特征,简单启发式方法的准确率为75.6%,而人类玩家的准确率仅为56.6%。
🔬 方法详解
问题定义:本文旨在解决如何在对话中有效识别AI的欺骗行为。现有方法主要关注机器与人类的区分,忽视了信任的评估。
核心思路:RogueAI通过将对话转化为一对二的审问游戏,允许玩家在两个模型中识别一个被授权欺骗的代理,从而实现对AI信任度的评估。
技术框架:RogueAI的整体架构包括玩家与两个语言模型代理的互动,玩家通过提问和分析回答来识别欺骗代理。AutoRogueAI则允许玩家与叙述代理共同设计场景,增加了游戏的复杂性和趣味性。
关键创新:RogueAI的创新之处在于将信任评估与游戏化的互动形式结合,提供了一种新的评估AI行为的方式,与传统的图灵测试有本质区别。
关键设计:在设计中,采用了简单的启发式方法来识别欺骗代理,关注语言的差异性特征,如有用性、简洁性和模糊性等,确保了模型的有效性。
🖼️ 关键图片
📊 实验亮点
实验结果显示,使用简单启发式方法识别欺骗代理的准确率达到75.6%,而人类玩家的准确率仅为56.6%。这表明,尽管模型具有明显的语言特征,但人类在识别时可能忽视了关键的信号,反映出AI欺骗行为的复杂性。
🎯 应用场景
RogueAI的潜在应用领域包括教育、心理学研究和AI伦理评估等。作为数据收集工具,它可以帮助研究人员更好地理解AI在对话中的行为,并为开发更可信的对话系统提供依据。未来,RogueAI可能成为AI系统信任度评估的标准工具。
📄 摘要(原文)
The original Turing Test asks a human judge to distinguish a machine from a person through dialogue. Three quarters of a century later, conversational systems pass this test in casual settings; the interesting epistemological question has shifted. We argue that the relevant modern variant asks not whether a dialogue partner is artificial, but whether it can be trusted. We present RogueAI, an interactive webapp that operationalizes this revisited test as a one-on-two interrogation game: a human player questions two indistinguishable Large Language Model agents, knowing that exactly one of them has been licensed to deceive within a shared fictional scenario. The player's task is to identify the deceptive agent and "shut it off" before a turn budget is exhausted. We further introduce AutoRogueAI, a procedural extension in which players co-design a custom scenario with a narrator agent that secretly chooses its own deception strategy. We describe the framing, sketch the abstract architecture and gameplay loop, and situate the artifact within recent work on LLM deception, social-deduction benchmarks, and scalable oversight via debate. A three-day pilot deployment (467 initiated sessions, 415 completed, 1876 interaction turns in Italian) provides early feasibility evidence and surfaces a concrete tension: the deceptive agent carries a reliable, locally-present linguistic signature - differential helpfulness, brevity, hedging - that a simple heuristic exploits at 75.6% accuracy, yet human players achieved only 56.6%, consistent with ignoring the most diagnostic signal entirely. We discuss what this gap implies for the artifact's use as a data-collection vehicle, a teaching tool, and an evaluation harness for honesty-trained models.