AI agents have been helping large and mid-sized banks and other financial institutions (FIs) — including FinTechs and payment providers — reduce the cost of their anti-money laundering (AML) operations for several years.
According to Workfusion, the technology is now reshaping how compliance teams handle everything from transaction monitoring to sanctions screening and adverse media reviews.
The bulk of the human resource investment in AML operations is concentrated in transaction monitoring (TM). Unlike other compliance processes such as customer due diligence, TM demands a more complex level of analysis. Transactions can introduce risk when they are deliberately structured to obscure the parties involved, the purpose behind the transfer, or the relationship between multiple transactions. This is particularly evident in layering — a technique in which bad actors break large financial transactions into smaller, individual ones to avoid scrutiny. TM complexity escalates quickly, especially when numerous parties are involved and money moves across multiple borders and jurisdictions.
The cost burden of TM is compounded by compliance leaders’ need to leave no stone unturned. Missing a genuine financial crime can invite very public regulatory fines, lasting brand damage, and costly remediation programmes spanning several years.Banks and other FIs are turning to WorkFusion AI agents to address the back end of the Detection > Alert Generation > Alert Review > Decisioning workflow. While the front end — Detection and Alert Generation — is already well served by existing tools, it is the labour-intensive Alert Review and Decisioning stages where resources have traditionally been concentrated. AI agents step into this gap, making human resources significantly more efficient.
Ongoing geopolitical instability continues to place fresh pressure on AML/CFT budgets. New and evolving sanctions regimes raise difficult questions: which individuals and entities represent sanctions evasion risk, and how can compliance teams realistically track them all? As with TM, banks tend to configure their sanctions screening systems conservatively, meaning nearly every transaction triggers an alert requiring review. False-positive rates of around 95% are again the norm. AI agents offer an effective solution — shifting compliance teams away from low-value alert triage so that investigators can focus on genuine risks, while helping FIs eliminate backlogs, accelerate payments, improve onboarding, and support business growth.
AI agents draw on a broad set of technologies — including machine learning (ML), natural language processing (NLP), optical character recognition (OCR), and generative AI — and bring all of them to bear in the adverse media review component of know your customer (KYC) operations.
KYC work encompasses suspicious activity detection, investigation, and reporting, and it is built around five core activities: gathering customer and third-party information; organising that information for operational teams; assessing what the data reveals; reasoning through relationships and behavioural patterns; and making judgement calls on whether a matter warrants escalation or can be closed.
AI agents perform all five of these activities autonomously. By accessing data from standard adverse media screening tools, they prioritise relevant news articles and other customer information, evaluating factors such as relevance, demographic data, and material significance. WorkFusion’s AI agent, named Evan, can conduct a comprehensive investigation into a person or entity using as little information as a name alone. Evan then highlights articles and data points that indicate risk, passing the results to a team member for review — along with a detailed, written, and auditable justification of its risk evaluation.
A typical adverse media search returns five or more results, each of which takes a human analyst between 10 and 20 minutes to read and assess. Evan reads and analyses the same material in just 2–3 minutes, reducing review times by 80–90%. The agent also routes the highest-risk cases to team members via human-in-the-loop technology, ensuring that human judgement remains central to the most critical decisions.
Ultimately, whether the task at hand is TM alert review, sanctions screening, or adverse media assessment, AI agents are proving to be valuable colleagues for AML analysts and investigators alike — taking on the tedious, budget-heavy work so that human teams can focus where they are needed most.
Copyright © 2026 FinTech Global
