AI is now a core part of the transaction monitoring conversation.
But our "AI adoption done right" webinar made one thing clear: the firms seeing real performance gains are not simply “adding AI” to existing processes.
In this session, Kathy Gormley, Head of Product (Transactions), and Lucie Novotna, Head of Solution Engineering (Transactions), explore why AI adoption in transaction monitoring is producing very different results across the industry.
In this session, Kathy Gormley, Head of Product (Transactions), and Lucie Novotna, Head of Solution Engineering (Transactions), explore why AI adoption in transaction monitoring is producing very different results across the industry.
- Some teams are using AI to move beyond rule-heavy alert generation, improve prioritization, and uncover suspicious behavour that traditional systems miss.
- Others are getting stuck with narrow automation projects, misleading efficiency metrics, or AI tools that optimise flawed workflows rather than improving detection outcomes.
Kathy and Lucie unpack where LLMs can add real value (for example, investigation support and SAR drafting) while explaining why detection itself requires models purpose-built for financial crime patterns, transaction behavior, and risk signals.
They also discuss why data quality matters, firms need to understand what data they have, where it is missing or inconsistent, and how those gaps affect model output, explainability, and auditability.
But don't let it stop you, as Lucie put it:
They also discuss why data quality matters, firms need to understand what data they have, where it is missing or inconsistent, and how those gaps affect model output, explainability, and auditability.
But don't let it stop you, as Lucie put it:
"You don’t need perfect data to start the journey; phased improvement is always possible."
Imperfect data can still be used if teams are transparent about its limits and improve coverage over time.
Watch the recording to hear more insights like this. Field-tested perspectives on AI-led transaction monitoring are at your fingertips.
Sign up to watch the recording now:
What you’ll learn:
- Where AI is delivering measurable value across detection, alert triage, investigation, and reporting.
- Why using LLMs for detection can miss the point (and where they are better placed in the workflow).
- How firms can start using AI even when their data is incomplete, fragmented, or still maturing.
- Why metrics like “time per alert” can be misleading when the highest-value investigations often require deeper analysis.
- How to avoid using AI to optimise broken processes instead of improving transaction monitoring outcomes.
- What separates teams seeing step-change improvements from those only achieving incremental gains.
- How governance, explainability, and human expertise fit into practical AI adoption.