Fresh data from Riskified, the digital fraud and risk intelligence company, shows that shoppers are getting increasingly more comfortable with agentic commerce. However, it has to be on “clearly defined” terms.

The latest results suggest consumers are becoming more comfortable with agentic commerce, but only on clearly defined terms.

In the company’s latest Agentic Commerce Pulse report, which polled more than 5,100 consumers globally, researchers learned that as AI agents like Muse and Instinct gain broad attention and adoption, “consumer comfort with the idea of using AI for shopping remains well ahead of actual adoption, and open-ended questions of control and governance still need to be addressed.”

“Half of all consumers surveyed by Riskified said they were more comfortable now than six months ago with the idea of an AI shopping agent completing a purchase without immediate approval, signaling growing openness to agent-led transactions,” the report’s authors said. “This is coming to pass as agents like Muse and Instinct are attracting a lot of focus and attention.”

The data showed there a “significant delta” between AI use for shopping discovery and comfort with the general idea of agentic commerce—and actually using an AI tool to complete a purchase. Of those surveyed, 80 percent said they have used AI for some portions of the shopping process. Seventy-four percent of consumers polled said they would allow agentic purchases in some form while 68 percent said they were likely to use AI for holiday shopping.

“Conversely, just 11 percent of consumers have completed a purchase inside an AI tool,” the report noted. “Meanwhile, openness toward AI declines the closer you get to the transaction: just 7 percent of people surveyed said they were comfortable letting AI buy something without their approval. The pattern suggests consumers are embracing AI as a shopping assistant faster than they are accepting it as an autonomous buyer.”

For Riskified, the findings suggest the companies are entering a “crucial moment” for the agentic market. “Existentially, Riskified’s Agentic Commerce Pulse points to larger issues of control and governance that still need to be addressed to truly clear the path for confident adoption and growth,” the report’s authors said.

“While consumers surveyed were increasingly comfortable with agentic commerce tools, the findings point to an emerging relationship of conditional autonomy, where consumers are willing to give AI greater authority if they establish the limits.”

The data showed that 43 percent of respondents surveyed by Riskified said they still want to approve every purchase before it happens, while 29 percent said they would allow an AI agent to purchase automatically within rules they set, “either requiring approval for exceptions or receiving notifications after purchases are made.”

AI governance and responsibility remains an open question. One-third of consumers polled said they would trust a general AI platform such as ChatGPT or Gemini to run a shopping agent, “compared with 17 percent who would trust a retailer or brand they already shop with, and 9 percent who would trust a bank, credit card company or digital wallet.”

This sentiment is presented despite security of payment credentials and AI choosing the wrong product being two of the biggest consumer concerns that were expressed. “When AI makes a mistake, the majority of consumers surveyed (51 percent) said the AI platform should be responsible for issuing refunds, even though Muse and Instinct both state in their terms and conditions that users are responsible for the actions taken by the agents,” the report stated.

Consumers also want AI to identify itself. Of those polled, 68 percent said they would always want to know when they are interacting with an AI agent rather than a human.

Jeff Otto, chief marketing officer at Riskified, said there’s a lot at stake. “Consumers are getting comfortable delegating shopping to AI, but they haven’t delegated the blame,” Otto said. “More than half expect the AI platform to refund a mistaken purchase, the platforms’ terms say the user is liable, and in practice, that dispute lands on the merchant.”

Otto said the agent could become a new refund-abuse vector. “Agentic commerce only scales if merchants can tell a real customer using a legitimate agent from a hijacked account or malicious automation,” he said. “That’s why Riskified is extending its identity risk intelligence to the person behind the agent, for the order and the service interactions that follow.”

The report’s authors said as agentic commerce expands, retailers and brands will need to distinguish trusted customers and legitimate AI agents from fraudsters attempting to exploit automated transactions. “Riskified helps merchants navigate that shift with AI-powered risk intelligence and identity-based insights to support safer commerce while reducing fraud and chargeback exposure,” the company said.