The Agent's Dilemma
We are currently in a bizarre transitional period. We have AI agents that are incredibly smart at processing text, but incredibly dumb at assessing source.
To a Large Language Model (LLM), a 2,000-word essay written by a subject matter expert with 20 years of experience looks mathematically very similar to a 2,000-word essay generated by a content farm yesterday. They are both just "tokens."
For decades, we relied on a gatekeeper to tell us who was real: Google. If you ranked high, you were "trustworthy."
But the search engine is dying. In a world where AI agents browse the web for us, we need a way to signal trust directly to the machine, without a middleman.
The Proposal: A Credibility Passport
I am working on a standard for this. I call it the Credibility Passport.
The concept is simple: Website owners embed structured data (like a digital passport) into their pages. This isn't for humans to read; it's for the AI agent scraping the page.
It answers the questions the AI should be asking, but currently can't.
Verifiability Over Claims
We have been down this road before. In the 90s, we had "Meta Keywords." You could just type <meta name="keywords" content="best website ever"> and search engines believed you. It became a spam fest.
The Credibility Passport is different because it relies on Verifiability, not Assertions.
- Don't tell me you're an expert. Show me the Provenance. Link to the Wayback Machine snapshot showing this domain has been active since 2012.
- Don't tell me you're active. Show me the Commit Log. Link to the structured history of updates to this specific page.
- Don't tell me you're unbiased. Show me the Disclosure. Tag the affiliate links and the methodology used to review the product.
Social Proof Without the Dox
One of the biggest issues with "Trust" online is the demand for PII (Personal Identifiable Information). "Upload your driver's license to get verified."
No.
We don't need to know where you live to know if you are credible.
- Trust Signal: "This author has published 500 articles on this domain over 10 years." (Verifiable).
- Trust Signal: "This domain is cited by these 50 reputable external sources." (Verifiable).
- Trust Signal: "This site has a published correction policy and a history of fixing errors." (Verifiable).
This protects the indie creator. You can be anonymous and still be credible, provided your work has a history.
Why This Works Better Than SEO
SEO became a game of "Ranking." It was adversarial. You were trying to trick the algorithm.
This new standard is Conversational. You are trying to help the Agent.
AI agents want to be accurate. They want to avoid hallucinations. By providing a Credibility Passport, you are effectively saying: "Here is the ground truth. Here is why you can cite me safely."
The "Anti-Slop" Defense
The internet is drowning in AI-generated "slop"—infinite pages of low-value text created to farm clicks.
Slop has no history. Slop has no methodology. Slop has no correction policy.
By standardizing these signals, we create a filter that slop cannot pass. We create a VIP lane for the humans (and the high-quality agents) who actually do the work.
The Protocol: We cannot wait for Google or OpenAI to build this for us. We need to build the standard ourselves. If you own your platform, you own your proof.
Get Started
I've built a free generator to help you create your own Credibility Passport. Fill out a simple form, and it outputs the JSON-LD and HTML code you can paste into your site's <head> section.
→ Generate Your Credibility Passport
The standard is open. The tools are free. If you believe in an internet where quality content can prove itself without gatekeepers, help spread the word.
