Major banks warn AI shopping agents pose fraud and scam risks

Six major global banks have issued a joint warning about the fraud and scam risks posed by AI agents that can make purchases on behalf of users, calling for the technology industry to adopt proactive safeguards.

Bank of America, Capital One, ASB Bank, Commonwealth Bank of Australia, ING Group, and NatWest Group published a principles paper on Tuesday titled "Building Trust in Agentic Commerce." The document outlines concerns that AI shopping assistants could expose consumers and merchants to new forms of financial harm.

The core concerns

The banks identify several specific risks in their paper. AI agents empowered to act autonomously could purchase incorrect items, overspend, or fall victim to scams and fraud. Consumers lack clarity about whether these AI systems will act in their best interests, and uncertainty exists about liability when things go wrong.

"As highly regulated financial entities, we are focused on managing risk effectively as emerging technologies arise," the paper states. "Consumers are unclear if AI agents will act in their interests. They are concerned that AI agents may buy the wrong thing or spend too much—or even worse, lose their money to scams and fraud."

The institutions also warn that malicious actors could exploit new attack vectors, including compromising or impersonating AI agents and merchants, and developing novel forms of social engineering. Some AI providers may engage in unsafe practices such as directly entering consumer card details into websites or prioritizing payment methods with weaker consumer protections.

Merchants face their own risks. The banks predict that AI-driven commerce could lead to increased credit card disputes and chargebacks when AI agents make unauthorized or incorrect purchases.

Five principles for safer AI commerce

The paper proposes five guiding principles: Transparency, Safety, Privacy & Data, Choice, and Interoperability. The banks want AI companies to prioritize consumer and merchant safety, obtain proper consent for data usage, and avoid locking users into restrictive ecosystems.

Why it matters

This coordinated statement from major financial institutions signals that AI commerce risks are moving from theoretical to immediate. As tech companies race to deploy AI agents capable of autonomous purchasing—Amazon, Meta, and others are actively building these systems—the entities responsible for processing payments and protecting consumers are demanding a voice in how the technology develops. The timing is particularly notable: the paper was released within 24 hours of reports about a zero-day vulnerability in Meta's Muse AI assistant and Amazon's decision to block Muse from its platform.

For businesses considering AI agent integration, the message is clear: financial institutions expect robust safety measures, not just innovation speed. The principles outlined could influence future regulations and industry standards for AI-powered commerce.

These details were first reported by Mashable.