Global financial messaging services provider Swift, together with member banks, has launched two AI-based pilots to combat cross-border payment fraud and save fraud-related costs.

Swift said that its AI governance framework enables accuracy, explainability, fairness, auditability, security, and privacy for every aspect of its AI applications.

The two pilots aim for responsible use of AI and are aligned with emerging global standards, such as ISO 42001, the NIST AI Risk Management Framework and the EU AI Act.

In the first pilot, Swift will upgrade its existing Payment Controls service using an AI model that will create a more nuanced and accurate picture of potential fraud activity.

The Payment Controls service helps financial institutions detect anomalies indicative of fraud using historical patterns of activity on the Swift network.

As part of the pilot, Swift will work with Payment Controls customers to refine the enhancement and will use the customers’ live traffic data, making the findings applicable in the real world.

Swift chief innovation officer Tom Zschach said: “AI has great potential to significantly reduce fraud in the financial industry. That’s an incredibly exciting prospect, but one that will require strong collaboration.

“Swift has a unique ability to bring financial organisations together to harness the benefits of AI in the interests of the industry, and we’re excited by the potential of both of these pilots to help further strengthen the cross-border payments ecosystem.”

The second pilot uses advanced AI technology to analyse data from different sources to strengthen the global financial ecosystem.

For the second pilot, Swift has convened 10 leading financial institutions, including BNY Mellon, Deutsche Bank, DNB, HSBC, Intesa Sanpaolo, and Standard Bank.

Swift and the banks will use secure data collaboration and federated learning technologies.

The pilot will use secure infrastructure that enables financial institutions to exchange relevant information with strong privacy-preserving controls.

Swift’s AI anomaly detection model will then gather insights and identify potential fraud patterns from a much larger dataset.

The two pilots could lead to the wider use of information sharing in fraud detection, based on its success in assessing cybersecurity threats, said the financial messaging services provider.