NOMINATION HIGHLIGHTS

The Problem Agent Checkpoint Was Built to Solve
AI agents are already transacting on business websites, accepting terms, initiating purchases, and accessing sensitive data, often without explicit human authorization. Across Vouched’s customer base, between 0.5% and 16% of all incoming traffic now comes from AI agents. Traditional cybersecurity tools were built to detect and protect against human threats. They are blind to the unique risks AI agents introduce: malicious actors using agents to bypass fraud controls at scale, agents accepting binding commitments without direct user approval, and organizations with no visibility into what agents are doing on their platforms.

What Agent Checkpoint Does
Agent Checkpoint is a full suite of agent identification and permissioning for website operators, giving organizations full visibility and control over AI agent activity through three core capabilities:

• Detect: See which AI agents are interacting with your platform, how they are classified, and the risk they present
• Enforce: Apply policies to allow, block, challenge, or redirect agent traffic based on risk
• Govern: Authorize verified agents under defined permissions, with explicit approval and full audit visibility

Agent Checkpoint also provides authentication to verify the identity of users and the agents acting on their behalf, delegation controls to define exactly what actions an agent is authorized to perform, legal authorization requirements before agents make binding commitments, instant revocation as risk conditions change, and complete audit trails for security, compliance, and oversight.

Built on Open Standards
Agent Checkpoint is built on KYA-OS, Vouched’s extension to Anthropic’s Model Context Protocol that defines identity and delegation for AI agents. With KYA-OS, agents can cryptographically prove who they represent and whether they have explicit permission to act.

Why It Matters
Agent Checkpoint is the first solution purpose-built for the agentic threat surface, giving organizations a practical, deployable path to governing AI agent activity today.