Finastra has linked up with Swiss fintech NetGuardians to launch an artificial intelligence (AI)-powered fraud detection tool for Swift messaging.

The tool, launched today at Sibos, is integrated with Finastra’s fusion total messaging system and uses a combination of AI and machine learning to proactively monitor for anomalies. It captures all network traffic and Swift MT 101, 103 and 202COV payments, analysing unusual user activity to ultimately suspend suspected fraudulent transactions before they leave the bank or corporate.

“The launch of our AI and machine learning-based fraud detection solution marks an important step forward in helping banks tackle the challenge of financial crime,” says Riteesh Singh, senior vice-president of financial messaging and services at Finastra.

Unlike rules-based solutions, which compare activity to specific parameters and flag those that fall outside the standard range, NetGuardian’s algorithms continuously learn about banks’ payment instruction patterns, allowing them to identify high-risk message anomalies without needing to configure any rules. The two companies say that this will result in improved detection and prevention rates, as well as a significant decrease in false positives and time spent on fraud investigation. According to Finastra, this new tool will also help banks conform to Swift’s mandatory customer security programme, established to support customers in the struggle against cyber-attacks.

“Fraud is a growing problem around the world and a big challenge for banks; and unfortunately, the Swift network has found itself a target in recent years. With regulators urging banks to be more vigilant to this threat, we have built a tool that empowers banks to proactively prevent fraudulent activity before it can occur and keeps them one step ahead in the fight against financial crime. We are already seeing great success with our technology,” says Joel Winteregg, co-founder and CEO at NetGuardians. He adds that NetGuardians’ customers have stopped more than US$1mn in fraudulent payments since January this year. “Now that we have integrated our tool with Finastra’s technology, this will only increase.”