<img height="1" width="1" style="display:none;" alt="" src="https://px.ads.linkedin.com/collect/?pid=1611884&amp;fmt=gif"> Skip to main content
NewVouched is live in Epic MyChartLearn More
  1. Solutions
  2. Use cases
  3. Govern AI agents

Use case: Govern AI agents

Know who every agent acts for.

AI agents now book and buy for people. Know Your Agent detects them on your site and APIs, enforces what each may do, and ties every action to the verified person who authorized it.

An AI agent governed by Know Your AgentA shopping agent acting for Maya arrives on your site and is classified as an AI agent. It presents a delegation from Maya, a verified person, is checked against your policy and scopes, and is verified and allowed, with an audit trail. AGENT REQUESTShopping agentActing for Maya R.1Arrives on your site, classified as an agent2Presents a delegation from a verified person3Checked against your policy and scopes4Allowed, with an audit trailRESULTVERIFIED · IN SCOPE

The problem

Agents look like users. Until they don’t.

Agent traffic is growing fast, and most systems can’t tell a helpful agent from a scraper, or know who either is acting for.

Agent traffic

2.4×

Growth in AI agent traffic in 30 days across Know Your Agent: Detect customers

AI fraud

$893M

Lost to complaints involving AI in 2025

Sources: Know Your Agent: Detect, June 2026 · FBI Internet Crime Report 2025

How it works

Detect, enforce, govern.

Three steps, from seeing agents to trusting the right ones.

  1. 1

    Detect

    Classify every session as human, AI agent, bot or incomplete, with a confidence score. Free.

  2. 2

    Enforce

    Write policy in plain language: allow, challenge, block or redirect, at your gateway or in your app.

  3. 3

    Govern

    Agents present a delegation from a verified person, with scopes and limits, proven on every request.

  4. 4

    Audit

    Every agent action is logged against the person who authorized it.

Compared

How agent controls hold up.

Blocking every bot loses good agents; allowing every bot lets in the rest.

Reset methodTells agents from humansLets good agents inTies actions to a verified personAudit trail
CAPTCHA and bot blockingPartlyBlocks many botsNoNoNo
Allow all trafficNoYesNoNo
API keysPartlyFor known integrationsPartlyIf they were issued oneNoTies to an app, not a personPartly
Vouched Know Your AgentYesHuman, agent, bot or incompleteYesPolicy decidesYesDelegations from verified peopleYesEvery request

The proof

Recognized for agent identity.

Independent recognition for Vouched’s agent authentication.

CB Insights

#1

Ranked first for agent authentication and authorization

Awards

4×

Cybersecurity Excellence Awards 2026 winner

Sources: CB Insights · Cybersecurity Excellence Awards 2026

Trust

Built for the security review.

Everything your security review asks for, from the SOC 2 report to the BAA, is on one page.

Visit the Trust Center
HIPAA and BAABuilt to support HIPAA compliance; BAA in place
NIST SP 800-63A IAL2Kantara-approved workflow
SOC 2 Type IIReport available under NDA
ISO 27001Information security management
Zero access to clinical recordsIdentity data only
Biometric privacyNotice and consent on every selfie

FAQ

AI agents, answered.

Straight answers about how Vouched helps you govern AI agents.

Contact us with a question
What does it mean to govern AI agents?

Seeing which agents reach your site and APIs, deciding what each may do, and tying every action to the verified person who authorized it.

How does Vouched detect AI agents?

Know Your Agent classifies every session as human, AI agent, bot or incomplete, with a confidence score from 0 to 100. Detection is free.

Can we allow some agents and block others?

Yes. You write policy in plain language, and each request gets an allow, challenge, block or redirect verdict, at your gateway or in your application.

How is an agent tied to a person?

The agent presents a delegation from a verified person, with scopes and limits, and proves it on every request.

Why not just block all bots?

Because helpful agents now act for real customers. Blocking them all loses business; letting them all in invites abuse.

How fast is agent traffic growing?

AI agent traffic grew 2.4 times in 30 days across Know Your Agent: Detect customers, according to June 2026 detection data.

Is every agent action auditable?

Yes. Every action is logged against the person who authorized it.

How do we get started?

Start with free detection in a few lines of code, then add enforcement and governance where you need them.