Rabobank Advances Fraud Prevention with RiskShield from INFORM
Oct 28, 2026
For 15 years, Rabobank and INFORM have worked closely together to continuously advance the bank’s fraud prevention capabilities. Building on RiskShield technology, Rabobank has expanded its fraud detection capabilities over the years with data driven analytics, advanced risk models and machine learning. Today, millions of transactions are analyzed every day, enabling the bank to identify fraud risks quickly and effectively. This enables Rabobank to detect fraud faster and more accurately and respond more effectively to new and evolving forms of fraud. The results include fewer false positives, lower fraud losses and fewer manual reviews.
With more than nine million customers in the Netherlands and a strong international presence, Rabobank processes large volumes of transactions every day that require rapid and reliable assessment. At the same time, fraud scenarios are becoming increasingly complex, spanning multiple channels, involving coordinated attacks, or emerging through subtle behavioral patterns. Instant payments add further pressure, as suspicious activities must be detected and assessed immediately.
Assessing fraud risks in milliseconds
Together with international AI and optimization specialist INFORM, Rabobank has continuously advanced its fraud prevention capabilities. Highly rule-based and predominantly manual processes have gradually evolved into a dynamic, real-time fraud detection environment. RiskShield evaluates a broad range of contextual information from internal systems and trusted external sources, combining it into a holistic risk assessment for each transaction. This enables the bank to detect suspicious activities and intervene before financial damage occurs. “We strive to handle every transaction within milliseconds and update our decision logic and fraud detection models almost daily,” says Anton Jan Cozijnsen, Head of External Fraud at Rabobank.
Rabobank uses the solution across areas including online banking, card fraud, and risk-based authentication. The combination of automated detection and targeted risk scoring reduces the manual review workload, speeds up alert handling, and allows teams to focus more on high-risk and relevant cases. INFORM’s Hybrid AI approach combines advanced detection scenarios with machine learning models to detect complex or previously unknown fraud patterns. Feedback from confirmed fraud cases, false positives, and previously undetected cases is continuously incorporated into the further development of the detection logic. For example, when a new form of subscription fraud emerged through deceptive online advertisements, Rabobank was able to implement additional detection logic within a short time and prevent most cases.
New risks require flexible fraud detection
As new payment methods and regulatory requirements emerge, fraud prevention continues to evolve. Rabobank therefore continuously adapts its environment to new fraud risks and transaction types and plans to further broaden the use of machine learning and advanced analytics. Its long-standing partnership with INFORM has evolved into a co-development model, with the bank’s operational experience continuously shaping the solution. “Since the beginning, we have closely collaborated with INFORM and co-developed the product,” says Cozijnsen. Rabobank now also uses RiskShield across additional use cases, including compliance monitoring for anti-money laundering, sanction screening, and market abuse.