As U.S. Regulators Warn Banks Over Agentic AI Risk, Nearly 60% of Americans Fear Losing Control Over Money While 44% Expect AI to Take Over Payments, New Data Shows

WHY THIS MATTERS: The data confirms a critical hurdle in the race toward fully autonomous financial services. While the industry is rushing to embed sophisticated generative AI and agentic commerce capabilities into payment flows, this research reveals a deep-seated disconnect with the end-user. The core issue is not technology adoption, but consumer trust. A vast majority of Americans fear losing control, opacity in decision-making, and misuse of financial data. This resistance is a vital early warning for developers and regulators alike. The trend toward instant payments and ‘invisible’ transactions (like agent-initiated payments and VRPs) is fundamentally undermined if the customer cannot see the point of consent or override a decision. For a sector built on accountability and controlled execution, this mandates an urgent pivot from prioritizing mere efficiency to engineering transparency and human-in-the-loop governance by design.

As U.S. regulators warn that genAI and agentic AI are becoming a growing risk issue for banks, and prepare further input on how artificial intelligence should be governed in financial services, new research from payment platform Aevi reveals:

• 59% of Americans fear losing control over how or when payments are made if AI is managing them 
• 54% worry they would be unable to stop or override AI decisions
• 53% are concerned about AI being hacked, while 50% fear misuse of their financial data
• 49% worry payments could happen without their clear consent, highlighting concerns around transparency and visibility
• Only 5% have no concerns, pointing to near-universal scepticism toward AI-led payments 

Aevi’s survey of over 3,000 adults exposes a clear trust gap as AI moves closer to real-world deployment in commerce, and while consumers recognise potential benefits such as faster checkout and fraud protection, most remain uncomfortable with systems that reduce visibility, control, or explicit consent in payments.

51% of Americans wouldn’t feel comfortable with anyone using AI to manage their money on their behalf, reinforcing resistance to delegation regardless of provider.

Crucially, 44% believe AI agents will manage everyday payments within the next decade, a growing disconnect between expectations and consumer readiness.

On what the findings mean for the future of payments, Aevi CEO Mike Camerling highlights a key distinction emerging from the research:

“Agentic commerce isn’t being rejected because it uses AI. It’s being rejected because it removes visibility, perceived control, and a clear moment of consent from payments. When people don’t understand what a system is doing or why a decision was made, trust erodes very quickly.

This distinction appears repeatedly throughout the findings, however, trust breaks down when AI begins to operate invisibly, continuously, or without clear consent.

In-person payments already solve many of the trust challenges around AI. There is a human interaction, a visible device, and a deliberate action to complete the transaction. AI’s role is not to replace that moment, but to enhance it by helping select the right offers, loyalty benefits, or payment methods at the point of sale.”

There is a critical tension for the payments industry: while many consumers expect AI-managed payments to become commonplace, they remain hesitant to embrace systems that reduce control, visibility, or clear consent.

FF NEWS TAKE: This research absolutely moves the needle by providing a human lens on the regulatory landscape. The industry’s push into agentic commerce cannot succeed without solving the control and visibility crisis. The key takeaway is that AI must augment the customer experience, not abstract it entirely. We should watch for regulators in major markets to use this sentiment to accelerate mandatory transparency requirements, pushing providers to build clear audit trails and easy-to-access decision overrides for AI-led payments.