AI & ML
WATCH
NOW
FintechOS: A new way forward
such as flagging transactions in countries where a user typically isn’ t present has now incorporated additional layers of intelligence.”
Alan provides an example of this evolution:“ If a customer’ s recent airport purchases indicate travel, the fraud risk may be adjusted accordingly. Companies also analyse broader patterns, identifying spikes in unusual transactions and assessing whether a specific transaction aligns with those trends.”
Mark notes that“ real-time predictive analytics has revolutionised fraud detection by enabling financial institutions to identify and prevent fraudulent activities instantly.” Traditional rule-based systems relied on static thresholds and historical data, often missing evolving fraud tactics.
“ AI-driven analytics continuously learn from real-time transactional and behavioural data that allow banks to detect issues within seconds,” Mark adds. This capability enables the identification of new fraud patterns that would be invisible to traditional methods.
These emerging patterns include synthetic identity fraud, where criminals combine real and fake information to create believable profiles and account takeover fraud, detected through deviations in login behaviour and transaction habits.
116 April 2025