Our Services Contact Us Newsletter

Experian Launches Agent Trust Framework for AI-Driven Commerce

Cartoon illustration of a friendly anthropomorphic AI microchip character

Experian has introduced Agent Trust, a human-to-agent binding framework designed to connect consumers with AI agents in a verifiable way for agentic commerce transactions.

The framework addresses a growing challenge in the digital identity space: as AI agents increasingly act on behalf of consumers to make purchases, book services, and conduct financial transactions, existing identity systems lack a reliable method for binding a human’s verified identity to the agent operating on their behalf.

Agent Trust is designed to verify that an AI agent has legitimate authorization from a specific consumer before completing a transaction. The system creates a cryptographic link between a person’s identity and the AI agent acting on their behalf, allowing merchants and service providers to confirm the agent’s authority in real time.

Experian, which acquired AtData earlier this year to strengthen its email identity capabilities, is positioning Agent Trust as infrastructure for the emerging agentic economy. The framework arrives as multiple organizations work to establish trust layers for autonomous AI transactions.

The FIDO Alliance has launched its own standards initiative for trusted AI agent authentication, while Identity Digital introduced DNSid to provide verifiable identity for AI agents. The convergence of these efforts signals that the identity industry views agent authentication as a critical near-term challenge.

Experian describes the agentic commerce market as one where machine-to-machine interactions initiate transactions without clear ownership of liability, creating uncertainty around agent accountability and consumer protection. Agent Trust is intended to resolve that ambiguity by establishing a verifiable chain of authorization from the consumer to the agent to the transaction.

Sources: BusinessWire

–

By the ID Tech Editorial Team