Global banking cooperative SWIFT plans to ring in 2025 by launching AI-enhanced fraud detection capabilities. The new function will give financial institutions more accurate insight into potentially fraudulent activities in real time.

Fraud attacks that enlist AI-generated deepfakes that impersonate high-level executives and AI-created synthetic identities are only rising. According to SWIFT, the global industry estimates the total cost of fraud in financial services to be $485 billion in 2023.

“Bad actors are using increasingly sophisticated tactics to commit financial crime, and the global financial industry needs to raise its defenses higher to ensure their customers can continue to transact globally with confidence,” said Jerome Piens, chief product officer at SWIFT.

The project relies on the billions of transactions that transverse SWIFT’s network annually as learning data for the AI engine. SWIFT strips identifiers from the data to protect privacy and replaces them with pseudonyms.

“The technology will identify suspicious patterns in real time, reducing fraud risk and ensuring a safer banking experience for clients,” added John McHugh, head of CIB Operations and Control at Standard Bank, in a prepared statement.

The new technology builds upon the cooperative’s Payment Control Service, which numerous small and midsize financial institutions use, and is part of SWIFT’s portfolio of AI innovations.

Participating banks in the Asia-Pacific region, Europe, the Middle East, and North America completed the pilot with SWIFT earlier this year.

Additionally, the organization is working with major financial institutions to explore other ways they could share data among institutions while maintaining data privacy. One such technology they are investigating is federated learning, a machine learning technique that trains models across multiple decentralized servers or devices without revealing proprietary data.

Data sharing also earned a panel discussion during SWIFT’s 2024 Sibos conference in Beijing. Representatives from Deutsche Bank, Intesa Sanpaolo, UniCredit and SWIFT recognized the benefits of data sharing but called for regulators to define a minimum level of data that could be shared to improve fraud detection, Finexta reported. Until such regulations are in place, financial institutions will likely hesitate to share transaction data.