Beyond the transaction: Four lessons on detecting financial crime hiding in plain sight
On 11 November, Fortify hosted How Fincrime Hides in Plain Sight, bringing together financial crime experts Ken Palla, Ciara Aitchison and James Nurse for a discussion moderated by Fortify CEO Karthik Tadinada.
As more customer interactions and financial transactions move online, institutions have access to a growing range of device, identity, behavioural and network intelligence. Much of it sits outside traditional Fraud and AML systems, despite its potential to improve detection and investigations.
The discussion explored how financial institutions can make better use of this information, which signals genuinely matter, and what it takes to turn intelligence into action. Here are four key learnings from the webinar.
1. Context can completely change a risk assessment
A transaction that looks routine in isolation can become suspicious when you consider the wider context.
Take two payments: £4,850 to a new beneficiary and £180 to an existing one. The larger payment might initially attract more scrutiny.
Now consider that the £180 payment follows a password reset, originates from an unfamiliar device and is associated with compromised credentials. Suddenly, the risk assessment looks very different.
The same principle applies to individual digital signals. A new device, unfamiliar IP address or unusual login isn’t necessarily evidence of fraud; its significance depends on the customer’s established behaviour and the other intelligence available.
Better fraud detection depends on understanding how these signals relate to one another, rather than assessing each in isolation.
2. Digital intelligence has value throughout the customer lifecycle
The same digital intelligence that helps detect fraud can also strengthen onboarding, AML transaction monitoring and investigations. Much of this intelligence is now available at relatively low cost through providers such as Google, Cloudflare and AWS.
Device history, identity information and behavioural patterns can help identify suspicious activity during account opening. Connections between accounts, devices and beneficiaries can reveal relationships that transaction monitoring alone might miss.
Once an investigation begins, this information can provide valuable context about a customer’s activity and potential links to other cases.
For financial institutions, the opportunity isn’t necessarily to invest in entirely new datasets. It’s to understand what useful intelligence is already accessible and how to put it to work across Fraud and AML operations.
3. AI should help investigators focus, not give them more data to review
The value of AI in financial crime investigations lies in identifying what matters – not presenting analysts with more information.
Investigators already face substantial workloads. Reviewing transaction histories, checking linked accounts, examining device activity and gathering information from different systems can consume time that would be better spent assessing risk.
AI agents can help by assembling relevant context, highlighting anomalies and drawing attention to connections that warrant further investigation.
This is where the distinction between data availability and decision support becomes important. An investigator doesn’t necessarily need to see 50 additional data points; they need to understand which ones are unusual, how they relate to the case and why they deserve attention.
Technology should support that judgement, with investigators retaining responsibility for the decisions they make.
4. Better collaboration requires more than sharing data
Connecting Fraud, AML and Cyber intelligence only delivers value when teams can act on what they learn.
Separate systems remain a significant obstacle. In our webinar audience poll, disconnected data was the most commonly identified barrier to cross-team intelligence sharing, followed by limited visibility into what other teams hold.
While making information accessible is an important first step, effective collaboration also requires clear ownership, practical handoffs and processes for translating new intelligence into improved controls.
For organisations looking to put these principles into practice, we’ve developed two resources:
Stronger Together: What Fraud and Cyber Can Learn from Each Other explores how the two disciplines can learn from one another to strengthen intelligence sharing, threat assessment and control effectiveness.
Shared Defence Companion Guide takes those principles further, with practical guidance for different financial services sectors.
Turning intelligence into better detection
Investigation findings should improve future detection, not remain buried in closed cases.
When investigators uncover a new connection, emerging threat or weakness in an existing control, that intelligence should feed back into the detection process. This creates opportunities to refine rules, test new approaches and respond more effectively as criminal behaviour changes.
Fortify’s AI-native Fraud and AML platform is built around this approach, bringing detection, investigations and rule development together. Governed AI agents help surface relevant intelligence and reduce manual work, while teams retain oversight of their controls and decisions.
For more practical examples and perspectives from the discussion, watch the full webinar, How Fincrime Hides in Plain Sight, on demand.
Sign up for the latest news and insights from Fortify

Turn risk into ROI
The Fortify team can help

Find out how we can support your prevention strategy

Related articles



.avif)

