AI-in-the-Loop：通过利用LLMs和联邦学习实现隐私保护的实时诈骗检测与对话诈骗
AI-in-the-Loop: Privacy Preserving Real-Time Scam Detection and Conversational Scambaiting by Leveraging LLMs and Federated Learning
AI总结：
本文提出了一种结合联邦学习和LLMs的AI-in-the-loop框架，用于实时诈骗检测与对话诈骗，实现了隐私保护与安全审查的统一。
摘要：
利用实时社会工程学的诈骗（如钓鱼、伪装和电话欺诈）仍然是数字平台上持续且不断演变的威胁。现有防御措施大多是被动的，在活跃互动期间提供有限的保护。我们提出了一种隐私保护的AI-in-the-loop框架，能够主动检测并干扰诈骗对话。该系统结合了指令调优的人工智能和一个安全意识强的效用函数，该函数在参与与危害最小化之间进行平衡，并利用联邦学习实现无需共享原始数据即可持续更新模型。实验评估显示，该系统能够产生流畅且吸引人的回应（词法复杂度低至22.3，参与度约为0.80），而人类研究证实其在现实性、安全性和有效性方面显著优于强基线。在联邦设置中，使用FedAvg训练的模型可维持高达30轮，同时保持高参与度（约0.80）、强相关性（约0.74）和低PII泄露（≤0.0085）。即使在差分隐私下，新颖性和安全性仍保持稳定，表明在不牺牲性能的情况下可以实现稳健的隐私保护。对守卫模型（LlamaGuard、LlamaGuard2/3、MD-Judge）的评估显示了一种简单模式：更严格的审查设置减少了暴露个人信息的可能性，但也会限制模型在对话中的参与程度。相反，更宽松的设置允许更长和更丰富的互动，这提高了诈骗检测能力，但以更高的隐私风险为代价。据我们所知，这是首个将实时诈骗诈骗、联邦隐私保护和校准的安全审查统一到主动防御范式中的框架。
英文摘要：
Scams exploiting real-time social engineering -- such as phishing, impersonation, and phone fraud -- remain a persistent and evolving threat across digital platforms. Existing defenses are largely reactive, offering limited protection during active interactions. We propose a privacy-preserving, AI-in-the-loop framework that proactively detects and disrupts scam conversations in real time. The system combines instruction-tuned artificial intelligence with a safety-aware utility function that balances engagement with harm minimization, and employs federated learning to enable continual model updates without raw data sharing. Experimental evaluations show that the system produces fluent and engaging responses (perplexity as low as 22.3, engagement $\approx$0.80), while human studies confirm significant gains in realism, safety, and effectiveness over strong baselines. In federated settings, models trained with FedAvg sustain up to 30 rounds while preserving high engagement ($\approx$0.80), strong relevance ($\approx$0.74), and low PII leakage ($\leq$0.0085). Even with differential privacy, novelty and safety remain stable, indicating that robust privacy can be achieved without sacrificing performance. The evaluation of guard models (LlamaGuard, LlamaGuard2/3, MD-Judge) shows a straightforward pattern: stricter moderation settings reduce the chance of exposing personal information, but they also limit how much the model engages in conversation. In contrast, more relaxed settings allow longer and richer interactions, which improve scam detection, but at the cost of higher privacy risk. To our knowledge, this is the first framework to unify real-time scam-baiting, federated privacy preservation, and calibrated safety moderation into a proactive defense paradigm.
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。
