Financial crime investigators are recruited for their analytical skills. Yet a large share of their working day is often spent on far less valuable tasks: trawling through systems, gathering data, compiling evidence, writing up findings and clearing high volumes of alerts that turn out to be benign.
This problem affects both Transaction Monitoring (TM) and Fraud teams. Analysts may have to move between many platforms before they hold enough information to reach a decision. Rising alert volumes, strict deadlines and more sophisticated criminal tactics are placing more strain on teams that are already stretched.
Artificial intelligence could reshape this operating model. Still, WorkFusion, a UiPath company, argues that simply swapping manual effort for Large Language Models (LLMs) is not the full answer.
Its AI Agent for TM and Fraud Investigations, Isaac, is built on what the company calls a “Goldilocks” principle. This means finding the right balance of AI, deterministic automation and human judgement, and coordinating all three across each investigation.
After a TM alert fires, an analyst may need to pull customer records, review transactions, research counterparties, check case history, compare expected and actual behaviour, map relationships and document everything.
An account takeover (ATO) alert in fraud can require evidence from detection engines, authentication services, card networks, digital identity platforms, behavioural intelligence tools and third-party sources. Isaac is designed to take on much of this “hunter-gatherer” work. In Fraud Alert Review, it connects with existing fraud engines and data sources, consolidates context, analyses transactions, drafts a narrative report and passes the finished work to investigators for human-in-the-loop review.
WorkFusion’s view is that each technology should do what it does best. LLMs add value when unstructured information needs interpreting, such as synthesising historic SAR narratives, reconciling previous dispositions against a new alert or making sense of messy counterparty data.
Transaction arithmetic is a different matter. Aggregation windows, velocity calculations, threshold proximity and peer comparisons should be calculated deterministically and fed to downstream AI as established facts, rather than generated by a model.
The result is that investigators move from gathering and writing to reviewing and deciding. This matters when regulatory deadlines meet unpredictable workloads. Volumes spike, staff take leave and complex cases overrun, which leads to the familiar “day-30 scramble.” AI Agents provide extra capacity by working alerts as soon as they arrive.
Early results look promising. One financial institution using Isaac for first-party fraud and ATO reviews cut manual research time by more than 70%, and its analysts handled two to three times their previous case volume. A regional bank applying Isaac to structuring alerts reported around 10% of alerts auto-grouped, roughly 60% auto-closed and one to three hours saved on cases that needed investigation.
The message for the industry is clear. The future of TM and Fraud is not 100% people, 100% LLMs or 100% rules. It is the right mix of all three.
Read the full Workfusion post here.
Copyright © 2026 RegTech Analyst
Copyright © 2026 RegTech Analyst
