AI agents are no longer just tools. They are becoming increasingly autonomous and capable participants in digital economies, capable of making decisions, executing contracts, and influencing public discourse. As these agents take on more responsibility, we must ensure they are trustworthy, accountable, and safe.

This proposal argues that “smart policy”—adaptive, verifiable, and enforceable by machines—is needed to govern these agents. Inspired by the financial concept of Know Your Customer (KYC), this proposal envisions a new framework: Know Your Agent (KYA). This involves giving every agent a verifiable digital identity, tracking their behavior, and embedding rules into their software code.

Instead of relying only on after-the-fact enforcement, smart policy should enable real-time compliance. This could help Canada manage risks, support innovation, and protect national interests as the Agentic Web grows.