发表机构
University of Surrey; Universitat Politècnica de Catalunya (UPC)（萨里大学; 加泰罗尼亚理工大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究通过角色感知的决策者-监督者框架，结合区块链审计，探讨AI监督在何种条件下改善网络欺诈决策，发现其价值取决于角色分配、校准和流量组成，而非单纯增加第二个模型。
AI 中文摘要
何时第二个人工智能（AI）组件能够改善主要网络欺诈决策，而非增加运营负担？我们通过一个具有区块链可审计性的角色感知决策者-监督者（DS）框架来研究这一问题，评估了四种结合集中式机器学习、联邦平均（FedAvg）训练的联邦元模型以及基础或量化低秩适配（QLoRA）大语言模型变体的方向性配置。分析比较了仅主要决策和监督决策，使用非硬欺诈性能、干预负担、条件校准、流量混合和审查容量敏感性、可靠性测试以及区块链生命周期控制。确定性硬门解决了89.994%的欺诈请求，将非硬群体作为主要AI决策设置。条件验证校准在部署回放中未产生一致的、可转移的监督优势。DS-3 QLoRA是干扰最小的监督配置，但其在F1和总错误方面仍不如其主要FedAvg阶段。在36种重新加权的流量混合中，监督仅在两种极端高欺诈场景下减少了DS-4 Base的总错误。区块链测试支持摘要验证、篡改检测、授权、一次性审查解决和最终确定后完整性，同时暴露了最终确定前单次写入的限制。结果表明，AI监督的价值取决于角色分配、校准、升级策略、流量组成和生命周期控制，而非仅取决于第二个模型的存在。
英文摘要
When does a second artificial intelligence (AI) component improve a primary network-fraud decision rather than add operational burden? We study this question through a role-aware Decider-Supervisor (DS) framework with blockchain auditability, evaluating four directional configurations that combine centralised machine learning, a Federated Averaging (FedAvg)-trained federated meta-model, and Base or Quantized Low-Rank Adaptation (QLoRA) large language model variants. The analysis compares primary-only and supervised decisions using non-hard fraud performance, intervention burden, conditional calibration, traffic-mix and Review-capacity sensitivity, dependability tests, and blockchain lifecycle controls. The deterministic hard gate resolves 89.994% of fraudulent requests, leaving the non-hard population as the main AI decision setting. Conditional validation calibration does not produce a consistently transferable supervisory advantage on deployment replay. DS-3 QLoRA is the least disruptive supervised configuration, but it still underperforms its primary FedAvg stage in F1 and total errors. Across 36 reweighted traffic mixtures, supervision reduces total errors only for DS-4 Base in two extreme high-fraud scenarios. Blockchain tests support digest verification, tamper detection, authorisation, single-use review resolution, and post-finalisation integrity, while exposing a pre-finalisation single-write limitation. The results show that the value of AI supervision depends on role assignment, calibration, escalation policy, traffic composition, and lifecycle controls rather than on the presence of a second model alone.
Comments27 pages, 10 figures, 13 tables
