Swift, a global leader in secure financial messaging, has announced a partnership with Google Cloud to develop advanced anti-fraud technologies utilizing AI and federated learning. This initiative aims to enhance fraud detection in cross-border payments.
In the first half of 2025, Swift plans to launch a sandbox using synthetic data to prototype learning from historical fraud incidents, collaborating with 12 global financial institutions. This effort builds on Swift's existing Payment Controls Service and follows a successful pilot across various regions.
Google Cloud will work alongside technology partners such as Rhino Health and Capgemini to create a secure solution that preserves privacy while enabling collaborative intelligence among financial institutions.
Rachel Levi, head of artificial intelligence at Swift, emphasized the organization's unique position to lead industry-wide efforts against fraud, while Andrea Gallego from Google Cloud highlighted the transformative potential of federated learning in fostering a resilient financial ecosystem.
The initiative seeks to address vulnerabilities in the payment lifecycle by promoting data sharing among institutions while adhering to privacy regulations. Federated learning allows model training without the need for sensitive data to leave individual institutions, significantly enhancing fraud detection capabilities.
This collaborative approach is expected to redefine how financial institutions combat fraud, reinforcing the integrity of the global financial system.