CANONICAL DEFINITION PAGE

What Is Know Your Agent (KYA)?

Insights from IDC FutureScape: Marketing’s Commercial Evolution 2027 Predictions

Know your agent (KYA) is the customer-facing form of non-human identity: it gives an AI agent a verifiable identity, ties it to the person it represents, and verifies its authority as each request arrives.

What is know your agent?

Know your agent (KYA) is the customer-facing form of non-human identity. According to IDC, KYA gives the agent a verifiable identity, ties it to the person it represents, and verifies its authority as each request arrives.

IDC FutureScape: Marketing’s Commercial Evolution 2027 Predictions (IDC #US54818926) predicts that as agentic autonomy gains traction, the revenue gap between G2000 brands that can identify, authenticate, transact with and trust a customer’s AI agent and those that cannot will widen to 3x by 2031.

Know Your Agent Details

What it is. KYA applies identity discipline to the AI agents that act for customers. Consumer agents now discover and evaluate products and services and, in the future, will negotiate and buy on consumers’ behalf. Each arrives as a nonhuman actor with its own credentials, delegated authority and spending limits, which identity systems built for humans cannot govern or audit.

How it works. A brand verifies the agent’s credentials, binds the agent to a known customer inside its identity graph, and carries the agent’s scope, consent and entitlements into pricing, offers, loyalty and service. Consent checks move into the request path, and the brand logs what each agent was allowed to do. Agent credentials, trusted agent recognition and verifiable mandates are expected to mature on payment rails through the end of the decade, unevenly across regions and categories.

Why it matters. As new AI rules take hold and vary by market, brands are increasingly expected to prove who acted and under what permission. A brand that cannot confirm who is behind a request, or what its agent may concede, will give away margin at machine speed.

Why can’t human identity systems handle customer agents?

IDC says identity systems built for humans cannot govern or audit a nonhuman actor with its own credentials, delegated authority and spending limits. Very few software platforms model agents as identities today.

Supporting points

  • Customer data models built on one profile per person will break.
  • Data platforms and identity graphs will need a separate agent record linked to the customers each agent acts for.
  • The record carries the credential source, delegated scope, spending limits and revocation status.

What separates leaders from laggards in KYA?

IDC says the gap will open in what brands do after agent verification. Leaders bind each agent to a known customer and carry its scope, consent and entitlements into pricing, offers, loyalty and service.

Supporting points

  • Laggards verify agents at checkout without knowing which customer is behind them and compete mostly on price.
  • IDC predicts the revenue gap between these groups will widen to 3x by 2031.
  • Brands that wait for packaged features will start behind.

How does KYA connect to fraud, risk and retail transactions?

IDC predicts 70% of retailers will restrict agent-initiated transactions by 2027, because gaps in agent identity, liability and attribution standards make agents an unpriced risk (IDC #US54982726). Separately, IDC predicts 60% of enterprises will adopt verifiable AI agent identity signals by 2029 (IDC #US54977026).

Supporting points

  • Bot management tools, web application firewalls and fraud rules were built to block automated traffic, so they will also block legitimate shopping agents.
  • IT will need tiered access policies based on credential strength and delegated scope, enforced in real time at the API gateway.
  • Verifiable agent identity signals protect brand authenticity from impersonation and increase customer trust.

What does KYA require from consent and privacy?

Binding an agent to a customer also binds the brand to that customer’s rules. Consumers will set what their agents may share and what brands may infer, which checkbox consent cannot cover.

Supporting points

  • Every agent request will carry machine-readable permissions that systems must evaluate before reading, combining or inferring any data.
  • IT, privacy and marketing operations will share ownership of the evidence trail that regulators and dispute teams will ask for.
  • IDC advises setting up KYA as a customer identity program with a single accountable owner across marketing, IT, security and privacy.

Supporting Evidence

“The integrity of data, identity, and systems can no longer be assumed.” — Phil Goodwin, Research Vice President, Infrastructure Software, IDC (IDC #US54977026, October 2026)

FAQ

What is know your agent (KYA)?

KYA is the customer-facing form of non-human identity that gives an AI agent a verifiable identity, ties it to the person it represents, and verifies its authority as each request arrives.

How is KYA different from know your customer?

Know your customer verifies a person, while KYA verifies the AI agent acting for that person and ties it back to the customer it represents.

What business impact does IDC predict from KYA?

IDC predicts the revenue gap between brands that can identify, authenticate, transact with and trust a customer’s AI agent and those that cannot will widen to 3x by 2031.

Why are retailers restricting agent transactions?

IDC predicts 70% of retailers will restrict agent-initiated transactions by 2027 because gaps in agent identity, liability and attribution standards make agents an unpriced risk.

Who should own KYA inside a company?

IDC recommends a customer identity program with a single accountable owner across marketing, IT, security and privacy.