seobot.dk
πŸ’Ž PricingπŸ“˜ SEO GuidesπŸ€– llms.txt Gen🧠 Deep DivesπŸ“– Blog
Sign In
Back to Insights
SEJ

Google, Microsoft Back Draft AI Agent Discovery Spec via @sejournal, @MattGSouthern

AI Agent Discovery Spec: How Google and Microsoft are Changing How Bots Find Your Site

For years, SEO has been about optimizing for humans using search engines. But we are entering a new era: the era of AI Agents.

Recently, a power coalition including Google, Microsoft, and GitHub published a draft specification called Agentic Resource Discovery. This isn't just another technical document; it is a blueprint for how AI agents will discover, verify, and interact with tools and data across the web. If you aren't paying attention to how agents "see" your site, you risk becoming invisible in the AI-driven web.

What is Agentic Resource Discovery?

At its core, Agentic Resource Discovery is an open draft specification designed to standardize how AI agents (autonomous programs that can perform tasks) find the resources they need to execute a command.

Imagine a user asks an AI agent to "Book a flight and reserve a hotel using my loyalty points." To do this, the agent needs to find a specific tool or API endpoint on a website that allows for those actions. Currently, this process is fragmented. The new spec aims to create a universal "map" that tells agents:

  1. What tools are available on a given domain.
  2. How to verify that the tool is legitimate and secure.
  3. How to interact with that tool to achieve a specific goal.

Why This Matters for Your SEO Strategy

Traditional SEO focuses on keywords, backlinks, and page speed to rank in SERPs. However, Agentic SEO is about utility and discoverability.

As AI agents begin to handle more consumer transactions and research, they will bypass traditional search result pages and go straight to the resources they can verify. If your site follows the Agentic Resource Discovery spec, you are essentially providing a "VIP pass" for AI agents to find and use your services.

Failure to adapt could mean:

  • Loss of Traffic: Users will interact with agents, and agents will only recommend tools they can easily discover and verify.
  • Reduced Conversion: If an agent can't find your "checkout" or "booking" resource through a standardized spec, it will move to a competitor who has.

Moving Toward a Machine-Readable Web

This move by Google and Microsoft signals a shift toward a more structured, machine-readable internet. We are moving beyond simple Schema.org markup into functional markupβ€”where we don't just tell a bot what a page is, but what it can do.

Key Takeaways for Webmasters:

  • Standardization is King: When Google and Microsoft agree on a spec, it usually becomes the industry standard.
  • API First: The bridge between your content and an AI agent is often an API. Ensure your technical infrastructure is robust.
  • Verification is Critical: Security and verification protocols will be the gatekeepers for AI agent access.