AI task spotlight | Edition no. 02: Prior SAR activity analysis
Every two weeks, we spotlight an AI task from Unit21's task library, something many compliance and fraud teams are configuring and running inside their workflows today.
This edition covers prior SAR activity analysis. It's a quieter problem than structuring, but a more pervasive one: investigators routinely open alerts without knowing the customer in front of them has already been reported. The task fixes that.
What Prior SAR Filings Are (and Why the Lookup Problem Matters)
When a financial institution files a SAR with FinCEN, that record doesn't expire. It becomes part of the investigative record for that customer, and any future alert involving the same customer has to be evaluated in light of it.
FinCEN makes this explicit in the context of continuing activity SARs. Under BSA guidance, if suspicious activity involving a previously reported subject continues, institutions are expected to file a continuing activity SAR, generally within 90 days of the prior filing. That obligation only applies if the investigator knows a prior SAR exists. If they don't surface that history before they begin their review, the program can miss the filing obligation entirely, or satisfy it inconsistently, depending on which investigator happens to catch it.
The issue isn't that institutions lack the records. Most programs have complete SAR filing histories in their system. The issue is retrieval. In most workflows, getting that history in front of an investigator requires them to leave the alert, search the filing database by subject, identify which records are relevant, verify they apply to the same customer, and manually note the count, status, and timing before returning to the case. That's the standard process. It runs in parallel with every alert. And it creates inconsistency: some investigators do it thoroughly, some skip it under caseload pressure, and the program has no systematic way to ensure it happens at all.
That's the bottleneck. Not the data, it's there. Not the analysis, once you have the history, interpreting it is straightforward. The bottleneck is that surfacing it has to be done manually every time, from scratch.
.png)
Introducing Unit21’s AI task: Prior SAR activity analysis
Prior SAR activity analysis: What it does
Prior SAR activity analysis automatically surfaces a customer's complete SAR filing history the moment an investigator opens a flagged alert. Instead of manually searching through historical records, investigators see a plain-language summary of everything already known about this customer's prior filings right inside their workflow, before they've reviewed a single transaction.
The agent automatically reviews:
- Number of prior SAR filings associated with the customer
- Current status of each filing (in review, submitted to FinCEN, continuing activity SAR, etc.)
- Recency of prior filings and time elapsed since the last report
- Whether prior filings share typologies or patterns with the current alert
Prior SAR activity analysis: What the agent outputs
- A plain-language SAR history summary surfaced directly in the investigator's alert view; no tab-switching, no manual lookup required
- A count of prior filings and the current status of each, so investigators know at a glance whether they're looking at a first-time flag or a repeat pattern
- Contextual signals that inform the investigation before it begins: "This customer has been reported to FinCEN twice in the past 18 months," delivered before a single transaction has been reviewed
Prior SAR activity analysis: Why this matters
Prior SAR history is one of the most consequential signals in AML investigation. A customer with no filing history who triggers a moderate alert has a different risk profile from a customer with two prior SARs that trigger the same alert. Investigators know this. The problem is that in most programs, the information isn't readily available when needed.
The gap isn't awareness. It's a process. When prior SAR lookup depends on an investigator's individual thoroughness, their time, their caseload, and their familiarity with the filing system, programs introduce inconsistency at the point where consistency matters most. Two analysts reviewing similar alerts will reach different decisions based on different information, not different judgments.
Prior SAR Activity Analysis closes that gap. It doesn't surface SAR history only when an investigator thinks to look for it; it surfaces automatically every time an alert opens. The investigator's job is to evaluate what the history means in context and make the call, not to assemble the record before beginning.
Because it's built on Unit21's Build Your Own Task framework, it can be configured to surface the filing fields your program needs, in the format your investigators expect to see. It runs in your existing alert workflow. Nothing new to learn.
About the AI task spotlight series
The AI Task Spotlight runs every two weeks. Each edition covers one task from Unit21's library: what it does, how it works, and who it's for. If a task is solving a real problem for one team, it can probably solve the same problem for yours.
Gal Perelman is the Product Marketing Lead at Unit21, where she spearheads go-to-market strategies for AI-driven risk and compliance solutions. With over a decade of experience in the fintech and fraud sectors, she has led high-impact launches for products like Watchlist Screening and AI Rule Recommendations.
Previously, Gal held marketing leadership roles at Design Pickle, Sightfull, and Lusha. She holds a Master’s degree from American University and a Bachelor’s from UCLA, and is dedicated to helping banks and fintechs navigate complex regulatory landscapes through innovative technology.
