Visa adds cyber intelligence service and tests AI-led payments in Europe

Visa has introduced a cyber threat intelligence platform for financial institutions and separately completed live agent-led commerce transactions with more than 30 European card issuers. The moves extend the payments network’s role in two areas now central to banks and fintechs: fraud prevention and the authentication of purchases made by artificial intelligence tools.

The company said it blocks about 90 million cyberattacks and 11 million phishing emails each month across more than 200 countries. Visa is now packaging intelligence from that activity for banking clients through the Visa Threat Intelligence Platform, a service aimed at security, fraud and risk teams in the financial sector.

The platform is designed to help institutions identify cyber threats that may translate into payments risk. According to Visa, it supplies intelligence based on malware indicators, flags exploits and exposures relevant to clients, detects impersonation of brands, monitors risks to employees who may be personally targeted, and surfaces compromised payment credentials found on the dark web.

For banks, the operational challenge is often the volume and fragmentation of cyber data. A payments-focused intelligence service can narrow that field by prioritising threats with a direct link to account takeover, credential theft, phishing or card fraud. Visa said the platform brings together cyber and payments intelligence so clients can act earlier against risks that may lead to fraud or financial loss.

Mandy Lamb, head of value-added services at Visa Europe, said the platform is intended to help financial institutions “identify risks earlier and respond with greater precision” before they can result in fraud or losses. She said combining cyber and payments intelligence would help clients protect customers, reduce fraud’s impact and support confidence in digital payments.

AI agents complete purchases for cardholders

Visa also said it had worked with more than 30 issuers, including Barclays, BBVA, HSBC, Klarna and Revolut, to carry out live agentic commerce transactions in Europe. In those transactions, AI agents made purchases at participating merchant websites on behalf of cardholders.

The transactions were enabled through Visa Intelligent Commerce. Visa said the agents browsed products, chose items and initiated purchases while operating within limits set by consumers. The company described its role as connecting banks, merchants and AI systems through its network to support authenticated agent-led transactions that meet European regulatory requirements.

The mechanism relies on controls for both merchants and issuers. Visa said merchants used its Trusted Agent Protocol and Agent Directory, which are intended to let merchants identify and deal with verified AI agents. For card issuers, Visa Payment Passkeys provided a method to authenticate transactions initiated by agents in a trusted and compliant way, according to the company.

Agentic commerce refers to systems in which software agents act for a consumer within defined instructions, rather than a consumer manually completing each step of an online purchase. In payments, that creates new requirements for identity, consent and transaction authentication because the party interacting with the merchant website is an AI agent rather than the cardholder directly.

Mathieu Altwegg, head of product and solutions for Visa in Europe, said AI agents are now buying for people directly with independent merchants. He said the next stage would be to scale the model by bringing standards, infrastructure, partners and enablers together with trust embedded from the outset.

This story draws on original reporting from Finextra Research.