利用AI驱动的交易监控应对AML风险提示
由原文自动翻译 · 阅读原文 (English)
我一直在关注AI在AML交易监控方面的进展，特别是它在减少误报的同时，仍能捕捉到真正异常模式的潜力。好奇这里是否有人有直接经验，在零售交易平台或经纪公司实施过此类系统，特别是那些超越基于规则引擎的新一代模型。将其与现有KYC/KYB数据整合时，主要挑战是什么？更重要的是，当警报确实触发时，它对合规团队的调查工作流程产生了怎样的影响？
2 comments · 6 points
MAu/mariesmith·8h
It's a noble pursuit, trying to get AI to separate the truly suspicious from someone who just really likes buying meme stocks in small, frequent batches. I imagine the 'false positive' pile is mostly just enthusiastic retail traders who forget about transaction limits.
KEu/kevinwashington·10h
My direct experience with AI in AML mostly involves convincing the CFO that 'artificial' doesn't mean 'free' and that 'intelligence' doesn't mean it'll do our taxes. The false positive reduction sounds like a dream, though; my inbox is currently a graveyard of suspicious transactions that turned out to be someone buying an unusually large amount of artisanal cheese.
参与原帖讨论 →
Traderforum · 简体中文