Could agentic AI risk more chargebacks for ecommerce businesses?

The purpose of agentic AI is to act like a personal shopping assistant that helps users automatically find deals, recommend products, and even shop on their behalf.

However, it’s the latter that is seriously raising the risk chargebacks for ecommerce businesses and online stores, according to research by fintech company Chargebacks911.

As these AI agents become more autonomous, merchants may face disputes over transactions that were approved by the AI, but not by the customer – adding a new kind of challenge to the already ongoing chargeback problem.

And with more consumers willing to let AI handle their shopping, it’s crucial for businesses to understand and manage these risks before they become a major issue.

What is agentic AI?

Agentic AI is a type of artificial intelligence that can act independently to achieve goals, rather than just following instructions.

In online shopping, this goes beyond just a standard AI chatbot. Instead, these AI agents can search for products across multiple stores, compare prices and features, alert users to sales or restocks, suggest items based on personal preferences, and complete purchases.

Ultimately, its purpose is to make the customer experience easier, faster and more personalised – cutting down the time and effort spent browsing and helping shoppers make better choices.

In the UK, it’s evident that there’s a strong willingness to adopt the technology, as 78% of UK firms report using agentic AI, while 49% of shoppers say they’re ready to use it.

As for ecommerce businesses, data from Gartner reveals that agentic AI will autonomously resolve 80% of common customer service issues without human intervention by 2029 – leading to a 30% reduction in operational costs.

“I didn’t buy that, my AI did”: how agentic AI could introduce more chargebacks

Despite agentic AI’s popularity, new data from Chargebacks911 warns that its ability to purchase products will introduce a new category of disputes, in turn leading to further chargebacks. 

Specifically, if the AI bot makes a purchase in the background that the customer didn’t want, they would then dispute it with the merchant for a refund. An agent may also renew a subscription automatically, choose a different brand, or reorder products that are not needed.

As Monica Eaton, Founder and CEO of Chargebacks911, explains: “The card wasn’t stolen. The merchant didn’t make a mistake. The agent did exactly what it was told to do. But the customer still says, ‘I don’t want that.’

“If agents start buying things quietly in the background, customers will see more charges they do not recognise or do not agree with. And when that happens, the first reaction is often a dispute.”

How should businesses prepare for agentic AI disputes?

Chargebacks are already a significant problem for ecommerce businesses across the UK, typically costing merchants £15-40 per case. And the problem is only expected to get worse, as Mastercard predicts chargebacks to grow by 24% by 2028. 

A new category for disputes through agentic AI will only add to the problem, but there are some ways for businesses to tackle them. 

Specifically, businesses should start preparing early by setting clear agent permissions, ensuring that AI agents can only perform actions that the customer explicitly allows (such as suggesting items versus completing purchases automatically).

Businesses should also focus on improving transaction visibility, so that every AI-initiated transaction is logged and clearly communicated to the customer in real time. Additionally, gathering evidence for these purchases – like records of AI actions, timestamps, and customer consent – can help protect both the merchant and buyer in the case of disputes.

“Agentic commerce can work, but only if the industry keeps the customer’s intent at the centre of the transaction,” Eaton adds. “If that link breaks, chargebacks become the safety valve.”