Legl’s law firm customers run hundreds or even thousands of watchlist checks every week. A lot of the screening flags potential matches with the individual in question. Until now, someone had to read through all the results to make informed decisions. Legl's new AI-assisted screening changes that.

There is a particular kind of dread familiar to anyone who has worked in compliance at a law firm. You run a client through a watchlist screen checking for sanctions, politically exposed persons, adverse media etc and the results come back. Not one hit. Not two. A cascade of potential matches, each one demanding attention, each one requiring you to dig through names, dates, countries, sources, and ask the same question over and over: is this actually who we’re dealing with?

Most of the time, the answer is no. One of the main challenges arises when the person you’re screening has a very common name. But once you start analysing the matches, it becomes quick obvious that the dates don't line up or the geography is wrong. But you still have to work through every result because there hasn’t been a sophisticated way to parse and categorise the information. Though uncommon, problematic matches like PEPs may be against the firm’s risk policy or trigger EDD workflows and it’s critical from a risk exposure point of view that these results are analysed carefully so the firm can make the right decisions.

This is the problem Legl has worked closely with our customers to tackle. The result is a new AI-assisted watchlist screening experience, available now to all Legl customers. It represents one of the most meaningful changes to compliance workflows giving MLROs, compliance teams and other stakeholders an intuitive, simple experience that is designed to enable faster and clearer decision-making.

The Scale of the Problem

Watchlist screening is a critical part of law firm AML processes. Under the Money Laundering Regulations, law firms must screen clients against sanctions lists, check for PEP status, and conduct ongoing monitoring for the duration of the relationship. Legl searches hundreds of lists in real time, covering sanctions regimes, PEP databases, fitness and probity records, and adverse media.

That comprehensiveness is the point. A thorough screen should cast a wide net. But unfortunately, wide nets catch a lot that isn't a fish.

Common names return dozens of potential matches. A client named David Jones might share a name with a sanctioned individual, a foreign politician, someone who appeared in a news story about financial crime fifteen years ago on the other side of the world. Every single one of those hits appears in the review queue. Every single one has to be evaluated. And in a busy law firm where fee-earners and compliance teams are juggling transactions and the thousand small fires of practice, compliance reviews pile up, get deferred, or get completed in a rush.

The fundamental challenge is the noise - the sheer volume of the alerts. The answer is not to automatically take decisions about who is or isn’t a match on the firm’s behalf as there are different risk policies and considerations that need to be taken into account for each of Legl’s customers. Instead, we make it easier than ever for a Legl user to understand who is very likely to be a match, and why. And the also who is unlikely to be a match, and why. This quality of parsing through information, grouping it, providing the signals for and against, and a clear summary of information enables teams to make fast decisions whilst maintaining robust risk-based processes.

The downstream consequences are significant if firms are not reviewing and clearly marking matches against the person in question. Unreviewed matches mean noisier ongoing monitoring. When a client's PEP or sanctions status hasn't been confirmed or ruled out, every subsequent monitoring alert carries the same fog of uncertainty. Teams end up reviewing the same underlying question repeatedly, from different angles, without ever reaching a clean answer.

What Legl Built

The new experience starts with a recognition that not all potential matches are created equal. Some results share a name and a country. Others share a name, a date of birth, a middle name, and a listed address in the same city. The starting point is that firms need to see clearly where the signals pointing to a likely match are very obvious, and why that is the case. So Legl’s new approach treats this very high likelihoods of matches differently.

When a watchlist check returns results, Legl's AI now analyses each hit against the known data for your client or the individual in question. It examines the signals that matter: names (including middle names, which are often the definitive differentiator), date of birth, location, associated entities, and the nature of the underlying listing. It produces a structured recommendation for each result, classifying it as high, medium, or low likelihood of being a genuine match.

High-likelihood matches receive elevated visual prominence. They surface first, clearly flagged, demanding the attention they deserve. Low-likelihood matches i.e. the ones where the name is similar but everything else is different, are grouped separately, with the option to dismiss them all in a single action.

That single action matters more than it might sound. What was previously a laborious process of opening each result, reading through the underlying data, manually recording a decision, and moving on is now a considered workflow. For the results that clearly don't represent the individual in question, one click removes them from the queue. For the results that need attention, the AI has already done the preliminary work surfacing the specific signals for and against the match, so the reviewer arrives at the decision point informed rather than starting from scratch.

The system also distinguishes what it found against what it didn't. If a match shares a name but lacks any date of birth correlation, if the geography is entirely different etc, that absence of corroborating evidence is itself informative. The AI surfaces both the signals in favour of a match and the signals against it, giving the reviewer a complete picture rather than a one-sided alert.

Built With Customers

This release did not emerge from a product team working in isolation. The design and iteration process was driven by direct engagement with some of Legl's largest and most compliance-sophisticated customers.

The feedback from customers has been overwhelmingly positive, particularly from Legl’s largest customers who are onboarding hundreds of new clients every week. Using AI-enabled approaches to filter, classify and rank results has reduced the manual review time, freeing up time for other work streams.

All customers have access to this new screening experience through an opt-in mechanism. If you’re a Legl user, you will see a banner clearly displayed inviting you to use the new AI-enabled screening experience.

What It Means for the People Who Do This Work

It is worth pausing on what watchlist review actually looks like from the inside. In some law firms, it falls to a small number of people, dedicated compliance officers, risk managers and AML analysts. But in some of Legl’s customers, fee-earners who are also trying to do their substantive legal work, are also responsible for working through this data. It is critical that firm’s MLROs and compliance teams are on top of genuine matches, that every PEP is on record and that any potential issues are understood against the backdrop of the firm’s risk policy.

That confidence is hard to maintain when the review process is defined by volume and manual effort. When there are fifty results to review and the tool doesn't help you distinguish the one that matters from the forty-nine that don't, every review carries the same cognitive weight. The work becomes numbing rather than analytical.

The new Legl experience is designed to restore analytical focus. By the time a reviewer reaches a high-likelihood match, the preliminary work has already been done to get the teams to faster decisions. The AI has cross-referenced what is known about the client with what the watchlist data contains. The reviewer's job is to exercise judgment on the result that has been surfaced, rather than wading through everything to find it.

For ongoing monitoring, the benefits compound. When the initial screening review is thorough and when genuine matches are confirmed and non-matches are cleanly dismissed, subsequent monitoring alerts are sharper. The system knows what it is watching for. The noise that had accumulated upstream no longer propagates downstream.

Available Now

The AI-assisted watchlist screening experience is live for all Legl customers. Opting in is straightforward from within your Legl account, and firms can move between the new and legacy views without losing any data or decisions.

For firms that haven't yet worked with Legl's screening tools, or for those considering a move from fragmented compliance stacks, the Legl team can provide a walkthrough on a live demo.

To enable AI-assisted watchlist screening in your Legl account, opt in by clicking on the banner on any screening results page. To speak with Legl's team about how this works in practice, get in touch.