Visa launches threat intelligence platform, tests AI agents

Visa introduced the Visa Threat Intelligence Platform to give banks payment-focused cyber threat data and tools to detect and respond to risks tied to payments. The company also completed live agentic commerce transactions in Europe with AI agents acting for more than 30 issuers, including Barclays, BBVA, HSBC, Klarna and Revolut.

The threat platform combines Visa’s cyber and payments data for use by security, fraud and risk teams at financial institutions. Features include malware-based indicators, alerts on exploits and exposures relevant to clients, detection of brand impersonation, monitoring of employee-targeted attacks and reports of compromised payment credentials found on the dark web. Visa reported that its network blocks about 90 million cyberattacks and 11 million phishing emails each month across more than 200 countries.

In separate trials, AI agents browsed merchant websites, selected items and initiated purchases on behalf of cardholders, operating within parameters set by consumers. The tests used Visa Intelligent Commerce to connect banks, merchants and AI systems and relied on a Trusted Agent Protocol and an Agent Directory to let merchants identify verified agents. Issuers authenticated agent-initiated payments using Visa Payment Passkeys.

Visa said the commerce trials were carried out in line with European regulatory requirements and aimed to test end-to-end interactions among banks, merchants and AI providers. Mandy Lamb, head of value-added services for Visa Europe, noted the platform allows institutions to identify risks earlier and respond more precisely to reduce fraud and financial loss.

Mathieu Altwegg, head of product and solutions for Visa in Europe, commented that agents are beginning to make purchases directly with independent merchants and described scaling the approach as the next phase, involving standards, infrastructure and partners with trust built in.

Visa presented the two developments as measures for banks to manage payment risk and to enable authenticated agent-driven commerce while meeting regulatory and authentication requirements.