Online shopping is going agentic. On Thursday, Experian announced the launch of Agent Trust, a human-to-agent binding service that binds verified human identities with the AI agents acting on their behalf. The release, launched in collaboration with Visa, Cloudflare and Skyfire, highlights that some of the biggest players in tech are looking to support the rollout of agentic commerce.

Experian’s Agent Trust will use a know your agent (KYA) framework to verify human and machine identities through tokenization, creating secure tokens that traverse with the agent and help retailers to trust the purchases made.

More broadly, the partnership highlights that enterprises and retailers may find agentic commerce to be a more significant channel for interacting with consumers in the future.

Just last year, Morgan Stanley released an estimate suggesting that agentic shoppers could represent $190 billion to $385 billion in U.S. commerce spending by 2030, capturing 10 to 20% of market share.

Welcome To The Age Of Agentic Commerce?

While agentic AI is still in its infancy, many consumers are already using chatbots to streamline their shopping. According to Adobe’s 2025 Holiday Shopping Recap , generative AI-driven retail traffic surged 693% in the 2025 holiday season, with one-third of shoppers using AI assistants.

However, despite this increase in demand, AI-driven payments remain relatively rudimentary. While startups like Perplexity and OpenAI have introduced AI commerce offerings to the market, with the former offering an “Instant Buy” option with PayPal.

These services have been limited in scope, with blocks on AI crawlers reducing the potential reach. At the same time, a lack of trust in AI agents, both among consumers and retailers, is slowing potential adoption.

“A lot of consumers are still a bit wary. They’re curious. They love the convenience, but they’re not sure whether they can trust this. And retailers and merchants feel the same way. How do they know that this agent that’s coming to their website to shop is legitimate? How do they know there’s a real human being behind this agent, and it’s not some sort of fraud or robotic attack?," Kathleen Peters, chief innovation officer at Experian told me in a video interview.

Peters says that the reason agentic commerce hasn’t got significant traction just yet is due to a lack of trust. From this perspective, robust verification of human and machine identities has the potential to increase consumer and merchant trust in this technology.

Building Trust In Agentic Commerce

Both Experian and Visa have been at the center of fraud checking and verifying human identities for decades, and are each well-placed to use this expertise to verify agent identities.

Another notable element of this partnership is the inclusion of Cloudflare, one of the largest content delivery networks (CDNs). Back in July 2025, Cloudflare announced it would be giving customer websites the option to block AI crawlers from traversing their content. Such blocks mean that neither AI search tools nor agents can access those sites.

Under this new partnership, Cloudflare customers will now be able to verify that an agent has been linked to a human identity under Experian’s verification system and allow access, while Visa’s trusted agent protocol can help merchants identify Visa-trusted agents.

If agentic commerce is going to take off, merchants need to be able to trust the mechanisms used to identify human and bot traffic. While there’s still a long way to go, verification of agents through highly-regulated financial providers like Experian and Visa has the potential to increase trust in agentic shopping long term.

“We have seen a lot of take up from consumers using AI-powered search to find the things that they want, but then they’re still going to traditional ways to make that purchase," Peters said. “What’s going to be different now is that as that trust builds, that these purchases can be safe, it’s going to change how we think about online shopping in general.”

Challenges Facing AI Shopping

In theory, agentic commerce has the potential to be a new channel for enterprises to interact with consumers, but there are concerns yet to be addressed.

“From my perspective, agentic commerce is evolving as an additional commerce channel, not a replacement for websites and apps. In the near term, most activity will stay anchored to familiar digital journeys, while agents increasingly assist with discovery and decisioning," Maanas Godugunur, senior director of fraud and identity at LexisNexis Risk Solutions, a data analytics provider, told me via email.

Godugunur suggests agentic commerce will move from direct interactions toward human-in-the-loop experiences, with agent-to-agent autonomy emerging later. He notes that multiple platforms and protocols are developing in parallel, and expects consolidation over the next few years with convergence toward a common protocol and trust/authorization patterns.

“The biggest blockers are customer adoption, trust, identity and proving that the individual behind this transaction has authorized the agent i.e. tying the agent to an individual. Agentic journeys can reduce visibility into the signals merchants rely on today to manage fraud and liability,” Godugunur said.

He also notes that KYA still remains unresolved, and that liability and dispute rules are still being established, with adoption not scaling until responsibility allocation, returns, chargebacks and authorization are workable and auditable across merchants, platforms and payment providers.

In this sense, agentic commerce still has a way to go in maturity before it sees widespread rollout, but the interest in AI shopping among consumers suggests there will be an appetite to experiment with these tools in the future.