Experian this week announced Agent Trust, a framework designed to create a verifiable link between consumers and the AI agents acting on their behalf in commercial transactions. The announcement, made on April 30, 2026, positions Experian alongside Visa, Cloudflare, and Skyfire in what the company describes as a growing ecosystem for agentic commerce infrastructure.

The launch addresses a structural problem that has grown alongside the proliferation of autonomous AI systems: when an AI agent places an order or authorizes a payment, a business on the receiving end has no reliable way to confirm whether a real, verified person originally authorized that action. Fraud, misrepresentation, and unauthorized transactions become plausible when no accountable identity sits behind an agent's actions.

The Know Your Agent framework

At the technical center of Agent Trust is a "Know Your Agent" (KYA) framework - a deliberate extension of the Know Your Customer (KYC) concept already embedded in financial compliance globally. According to the company's announcement, the system creates what Experian calls Human-to-Agent Binding: a secure, persistent link that connects a verified individual, their registered device, and the AI agents they authorize.

The binding produces an Agent Trust Token, issued by Experian in real time, that validates both identity and transaction fraud risk at the moment of a transaction. These tokens are not static credentials. They sit within a broader Experian Agent Registry, which maintains a dynamic trust score for each registered AI agent. That score shifts over time based on behavioral signals and other risk indicators - so an agent that begins deviating from patterns associated with its registered human can see its trust rating fall before a transaction completes.

The system is designed to be platform-agnostic. According to the announcement, it is intended to work with existing payment infrastructure rather than requiring merchants or payment networks to rebuild their systems. Integration with Visa's Trusted Agent Protocol is a stated design goal. The framework also maintains what the company describes as a continuous, auditable history linking who the user is to how the agent behaves over time.

Experian stated that its existing fraud prevention capabilities already help clients avoid an estimated $15-19 billion in fraud losses annually. That figure provides context for the scale at which the company believes agent-based fraud risks could emerge as autonomous purchasing systems spread.

Partners and their roles

The announcement names three external collaborators - Visa, Cloudflare, and Skyfire - each contributing distinct pieces of the trust infrastructure.

Visa brings the Trusted Agent Protocol, a standard the payment network developed to help merchants identify and verify AI agents through Visa's existing network. The protocol is designed to confirm that an agent carries legitimate authorization to act on behalf of a consumer. According to Rubail Birwadker, SVP and Head of Growth Products and Partnerships at Visa, "Visa has spent decades earning trust across global commerce, which matters even more as AI becomes part of how transactions happen." Birwadker described Visa Intelligent Commerce and Trusted Agent Protocol as providing "the secure foundation for agentic commerce experiences at global scale."

Cloudflare's role is at the network level. The company enforces the trust layer at the edge of the internet - infrastructure it already uses to power and protect approximately 20% of global internet traffic. Cloudflare also operates what it describes as a next-generation AI stack for building, securing, and scaling agents. According to Stephanie Cohen, Chief Strategy Officer at Cloudflare, "the rise of AI agents represents one of the most significant shifts in the history of digital commerce, but it can only succeed if the underlying infrastructure is rooted in trust." Cohen described the combination of Cloudflare's network footprint with Experian's identity expertise as giving businesses "the tools to participate in agentic commerce with confidence."

Skyfire contributes an open, standardized method for packaging and exchanging agent-related information across platforms - addressing the interoperability challenge that arises when AI agents from different systems and providers need to interact with merchants and payment networks built on different standards. According to Amir Sarhangi, CEO and co-founder of Skyfire, "agentic commerce only works if merchants can confidently understand who they are transacting with, and if agents can pay as reliably as people do." Sarhangi described Skyfire's work on the Know Your Agent (KYAPay) protocol as enabling "a seamless, interoperable trust layer that brings together identity and payments to unlock the full potential of autonomous transactions."

Beyond these three, Visa Intelligent Commerce adds a further layer: network tokenization for secure AI commerce transactions. Network tokenization replaces raw card data with tokens tied to specific use cases, limiting the damage potential of any single credential compromise.

How a transaction flows

Experian's announcement includes a practical example that illustrates how the components work together. A consumer asks an AI agent to find the best noise-cancelling headset ahead of a trip. The agent evaluates products based on the consumer's stated preferences, surfaces a recommendation - in the example, a Bose headphone - and prepares the transaction for approval. When the consumer authorizes the purchase, Human-to-Agent Binding confirms that the agent is acting on behalf of a verified, real individual. The Agent Trust Token validates identity and fraud risk in real time. The merchant receives a cryptographically grounded confirmation rather than a bare instruction from an anonymous automated system.

This flow depends on a prior enrollment step that ties the consumer's identity, device, and authorized agents together within Experian's registry. The token issued at purchase time reflects not just the consumer's verified identity but also the current behavioral risk profile of the specific agent involved.

Why this matters for the marketing and commerce ecosystem

The marketing community has watched the agentic commerce debate intensify since mid-2025, when structural questions emerged about whether AI shopping agents could scale without fracturing existing retail relationships. The trust and identity layer has been a persistent gap in the emerging architecture.

Cloudflare and the payment networks began working on this problem directly. Cloudflare partnered with Visa and Mastercard in October 2025 to develop security protocols for AI agent shopping - including the Trusted Agent Protocol and Agent Pay - both built on Cloudflare's Web Bot Auth foundation, using HTTP Message Signatures with public key cryptography. That work established the authentication layer at the merchant-facing end. Experian's Agent Trust adds the consumer-facing identity binding that was not addressed by those protocols alone.

UK research published in November 2025 found that 85% of consumers planning to use AI for Christmas shopping would trust agents to place orders and execute payments. McKinsey projections cited in that research estimated agentic commerce could orchestrate between $900 billion and $1 trillion in US B2C retail revenue by 2030, with global figures reaching $3-5 trillion. Against those projections, the identity problem Experian is targeting becomes commercially significant at a scale that justifies infrastructure investment.

Google launched the Universal Commerce Protocol with major retailers in January 2026, establishing technical standards for AI agents to execute purchases. Mastercard introduced Agent Pay the same month. These developments built the transaction-side rails. Experian's announcement today completes a different segment: the identity and accountability side.

HUMAN Security expanded its agentic visibility tools in April 2026, noting that automation is growing eight times faster than human web traffic. The company processes approximately 20 trillion weekly transactions across 3 billion devices, and has noted a 7,000% increase in agent-generated interactions. At that rate of growth, unverified agent transactions become an acute operational and fraud problem rather than a theoretical future concern.

For advertisers and marketers, the implications extend beyond fraud prevention. Attribution models, audience targeting, and conversion measurement all depend on understanding whether an action was taken by a human or an automated system acting on their behalf. If the identity layer for agentic transactions is not established early, the measurement infrastructure built around human behavior becomes unreliable. The KYA framework, if widely adopted, would give marketing systems a verifiable signal about the nature of a transaction's origin.

Amazon formalized its own AI agent policy in February 2026, requiring AI agents accessing its marketplace to identify themselves under a new Business Solutions Agreement update effective March 4, 2026. That policy took a governance-and-compliance approach. Experian's framework takes a cryptographic and identity-binding approach - the two are complementary but stem from different starting points.

Experian's positioning

Experian is a FTSE 100 company listed on the London Stock Exchange under the ticker EXPN. The company employs 25,200 people across 33 countries, with corporate headquarters in Dublin, Ireland. Its existing business spans financial services, healthcare, automotive, agrifinance, and insurance, with identity verification and fraud prevention as core capabilities across all of them.

The company frames Agent Trust as a "natural extension" of its existing verification role, according to Kathleen Peters, Chief Innovation Officer at Experian. "Agentic commerce will not scale without trust," Peters said. "What's required is verifying the agent, the human behind it, and their intent to purchase. This is a natural extension of Experian's verification role in the ecosystem. We already help define trust in financial transactions; now we're bringing that same leadership to agentic commerce."

Peters' framing reflects a broader pattern in how established identity and financial infrastructure companies are entering the agentic commerce space. Rather than building new AI consumer applications, they are inserting themselves into the trust and authentication layer that those applications will require to function at commercial scale.

The announcement does not include details on pricing, deployment timelines, or which specific merchant or AI platform integrations are live at launch. The announcement states that Agent Trust services are "platform-agnostic and built to scale with the evolving agent ecosystem."

Timeline

Summary

Who: Experian, a FTSE 100 global data and technology company listed on the London Stock Exchange (EXPN), headquartered in Dublin, Ireland, with 25,200 employees across 33 countries. Ecosystem partners include Visa, Cloudflare, and Skyfire.

What: The launch of Experian Agent Trust, a "Know Your Agent" (KYA) identity framework that creates Human-to-Agent Binding - a secure, persistent link between verified consumers, their devices, and the AI agents they authorize. The system issues real-time Agent Trust Tokens validating identity and fraud risk, maintains an Agent Registry with dynamic trust scoring, and is designed to integrate with existing payment infrastructure including Visa's Trusted Agent Protocol.

When: Announced on April 30, 2026.

Where: Experian is headquartered in Costa Mesa, California for its North American operations, with corporate headquarters in Dublin, Ireland. The framework is designed as a platform-agnostic, global service.

Why: As AI agents begin to search and transact autonomously, businesses face the problem of confirming whether a real, verified person authorized any given action. Without a binding between agents and verified human identities, autonomous commerce introduces fraud, misrepresentation, and unauthorized transaction risks. Experian's framework addresses the identity and accountability gap that existing payment authentication protocols - focused on the merchant side - did not cover.

Share this article
The link has been copied!