KYA Doesn’t Solve Identity. It Inherits It.
AI Agents Are Becoming Economic Actors
AI agents are no longer just answering questions. They’re signing up for accounts, negotiating terms, and completing payments — all without a human clicking “confirm.” This adds a great deal of convenience to our lives, but it also raises a real question: when a request comes from an agent instead of a person, who’s actually behind it? That’s the question behind Know Your Agent (KYA) — giving enterprises visibility into who’s on the other end, and building that verification directly into application delivery infrastructure, is key.
This isn’t a fringe use case anymore.
It’s the leading edge of a shift that the industry doesn’t fully have a framework for yet. Nobody’s fully solved this for the markets where identity changes fastest. What’s needed is a persistent identity layer that keeps the real person behind a credential resolvable and current, not just verified once at signup. That’s the exact gap we built the Living Identity® Graph to close.
The Gap Nobody’s Naming
KYC verified the human. It was never built for a counterparty that isn’t a person at all — software, acting on behalf of a real person.
That’s the gap KYA (Know Your Agent) is trying to close. Some emerging protocols now issue agents a signed credential — a token disclosing verified information about who controls it. When a service requires proof of identity, the agent presents the token instead of a human filling out a form.
It’s a step forward but it still falls short. A valid credential establishes trust in the credential and the claims it carries. But maintaining confidence in the underlying human identity as real-world data changes requires a deeper identity layer.
An Agent Is a Proxy, Not an Identity
An AI agent has no identity of its own. It’s a proxy, acting on behalf of a real person. Validating the agent and its credential is one layer of trust. Maintaining confidence in the underlying human identity across systems and over time is another.
Governance frameworks are increasingly emphasizing accountability, transparency, traceability, and human oversight around AI systems. Emerging frameworks, including the EU AI Act, the UK’s Digital Verification Services Trust Framework, and NIST’s AI Risk Management Framework, reinforce that broader direction toward trustworthy and accountable AI.
The risk of skipping that step isn’t theoretical. Experian’s 2026 Future of Fraud Forecast points to growing uncertainty around agent ownership and liability as machine-to-machine transactions scale, with a majority of companies already reporting rising fraud losses year over year. And the geography of that risk matters: identity fraud rates fell in Europe and North America last year, while rising 9.3% in Africa, 16.4% in APAC, and nearly 20% in the Middle East. These are high-growth markets where agent-driven commerce is beginning to accelerate — and where maintaining identity continuity can be especially challenging.
The Industry Is Already Building the Second Layer of Identity
We’re already seeing a second layer emerge on top of Know Your Agent trust frameworks — one that takes the signed JSON Web Tokens carrying verified identity claims about the agent and its human authorizer, and enriches those claims against additional identity data with real-time risk intelligence.
That layered architecture matters: protocol-level identity and payment tokenization on one side, consumer identity enrichment and fraud scoring on the other. The credential and the deeper identity intelligence solve related, but distinct, trust problems.
The open question is what maintaining confidence in the underlying human identity actually requires as agent-driven commerce expands into markets with very different data conditions. Identity verification built around dense, stable historical records — the kind that underpins much of the credit bureau model — works exceptionally well where those records exist and stay current. High-growth markets present a different condition: consumers switch mobile carriers, change addresses, and move in and out of formal banking relationships more frequently, which describes much of LATAM, MEA/MENA, SEA, and Africa.
As such, these markets can benefit from identity infrastructure designed to maintain continuity as underlying data changes. Closing that gap requires interoperable agent credentialing to work alongside a persistent identity graph designed for markets where consumer identity data changes quickly.
Resolving the Principal Behind the Agent
KYA establishes trust in the agent and its authorization. Living Identity® addresses the persistence and continuity of the real-world identity behind it as data changes.
This is where the problem stops being about agent engineering and starts being about identity infrastructure. The credential can establish verified claims about the agent and its human authorizer. The harder problem is ensuring that identity remains current, consistent, and resolvable over time — something a single point-in-time check cannot address.
That requires a persistent, deterministic identity layer capable of maintaining continuity behind the credential. That’s the layer we built the Living Identity® Graph to be: a persistent, deterministic identity spine spanning 2 billion-plus verified profiles across 30 high-growth markets, with 95% adult population coverage across LATAM, MEA/MENA, SEA, and Africa.
In practice, maintaining confidence in the identity behind an agent draws on three distinct Living Identity® capabilities:
Graph ID analyzes the identity’s relationships and network structure — revealing shared attributes, connections, community patterns, and anomalous network behavior that provide additional context around the identity behind the agent.
Match ID confirms that the specific fields the agent’s credential discloses — name, phone, address — actually match the verified profile, rather than simply resembling it.
Enrich ID provides current identity attributes to strengthen and complete the verified profile, helping prevent the identity claim from becoming outdated as phone numbers, addresses, and other details change.
The Agent Economy Still Runs on Human Identity
Agents will keep getting faster, more autonomous, and more capable of acting without a human in the loop. None of that changes what ultimately sits behind the transaction: a real person. KYA is a meaningful step toward establishing trusted identity and authorization for agents. But as agentic commerce scales, maintaining confidence in the human identity behind those credentials becomes a separate and persistent identity challenge.
The credential establishes trust in the agent. The identity layer has to ensure that the real-world identity behind it continues to hold.
Explore how the Living Identity® Graph resolves the real-world identity behind the claim across 30 high-growth markets.
