An Open Standard That Gives AI Systems a Systematic View of Your Business
AI systems now answer questions about your business. The problem is that they often get it wrong.
Consider a typical situation. Brand products, services, technology, locations, leadership, and relationships are spread across multiple pages. The AI model retrieves the pieces from those pages, puts them together probabilistically, and generates an answer. The result is often brand names associated with ideas, fictitious executives, misquoted skills, and weak or non-existent attribution.
This is not a failure of AI models. It is a failure of the medium itself. We built the web on pages, links, and prose. AI retrieval systems require something quite different: a systematic layer of explanation and evidence.
Suggestion: EntityMap
EntityMap recently entered public consultation. It’s a new open standard that gives organizations a way to publish a single structured file. This file declares what the organization knows, maps how its key entities relate to each other, and links every claim back to the source evidence.
The consultation runs until 30 June 2026, with the official launch scheduled for 1 July. Over the next 33 days, the project is actively seeking usability feedback, technical critiques, and real-world testing from developers, SEO experts, publishers, structured data experts, and anyone who builds or relies on AI retrieval systems.
Where EntityMap Lives in Standard Landscapes
EntityMap does not replace existing web standards. It fills a gap that sitemap.xml and schema.org were never designed to address.
Sitemap.xml tells crawlers what pages are on the website. Schema.org defines what appears on individual pages. EntityMap tells AI systems what an entity is, what it knows, and how that information connects to the rest of the website.
This distinction is important. Consider a health care organization that publishes medical procedures. With schema.org, you can define a single page. With EntityMap, you can say the following: “Here are our main clinical areas. Here are the relationships between them. Here is the peer-reviewed evidence that supports each claim. Here is where that evidence lives on our site.” An AI program that reads that file gets a structured view of the facility’s information rather than reconstructing it from page fragments.
Or, consider a SaaS company concerned about how AI programs define its product. EntityMap allows a company to announce: “We offer feature X. We are different from our competitors in Y. Here is the proof: link to documentation, link to example, link to comparison page.” A company should not rely on LLM to see the difference in widespread web content.
The same logic applies to publishers protecting attribution, law firms clarifying professional boundaries, financial services firms navigating regulatory nuance, and brands concerned about AI misrepresentation.
How EntityMap works
An EntityMap is a JSON file published to a predefined location in the domain. It consists of three important elements.
Businesses named things that the organization includes: products, services, people, concepts, places, regulations, technical areas.
The relationship map how those entities interact. Examples: “this product develops this effect,” “this person leads this group,” “this rule governs this service.”
Parts of evidence they support episodes from websites, linked to their source URL.
Each episode carries attribute metadata: publisher name, source page, retrieval timestamp. This metadata survives the extraction, integration and storage of vector information. When an AI system generates a response using your content, the chain of evidence remains intact.
The specification is intentionally small. The consensus floor contains about 12 fields that are required for all three items. Everything else is optional enrichment: custom predictions, cross-shard tuning, validation status announcements, log tracking.
Who Should Pay Attention
When building Retrieval Augmented Generation (RAG) systems, cleaner source data means better chains of reasoning and fewer observations.
If you are an SEO expert, this represents a new lever for AI visibility. It works with traditional content and linking techniques rather than replacing them.
If you are a publisher, this is a way to publicize what you know and maintain attribution as your content is classified across AI platforms.
If you are concerned about how AI programs represent your organization, this is a tool to ensure control.
Standard published under CC BY 4.0. No vendor lock-in, no registration, no proprietary software required. Public offering is open. The source code, specifications, and validation tools are all available on GitHub.
What the Project Needs From You
Consultation time is not for celebration. The project team is actively seeking specific types of feedback.
Response to technology use: Have you ever tried to create an EntityMap for your site or product? What is broken? What felt difficult in the performance?
Validate usage: Does this really solve the problem you are facing? Is it missing something important in your domain or industry?
Criticism prediction: The standard defines 24 core predicates (IMPROVES, DEEPENDS_ON, MEASURES, and others). Are these the right semantic summaries for your work? Should we add or remove from this list?
Integration ideas: Do you build a generator? Confirmation? EntityMaps management dashboard? The project wants to know what tools you have in mind.
Industry specific applications: If you work in health care, finance, education, law, or any other vertical, what would an EntityMap profile for your industry look like?
The specification is available at entitymap.org/spec/v1.0. The validator is live at entitymap.org/validate. The public forum and GitHub repository are at github.com/entitymap.
Participants are invited to review the specification, test implementation, raise issues, suggest improvements, and participate in the discussion before June 30, 2026.
Key Context: This Is Really Open
This is a standards proposal from within the search and AI community. RV Guha, one of the founders of schema.org, reviewed the project and gave it his approval.
The consultation is really open. The first phase focuses on technical review and early implementation. Wider adoption, sector-specific applications and research into the wider impact of the standard will follow after the consultation closes.
Why This Moment Matters
If you’ve spent the last few years watching AI programs misrepresent your work, your clients’ work, or your organization’s expertise, now is your time to plan how that changes.
The bar for entry is low. You need to review the specification, check it against the real problem you care about, and tell the project what you found. That feedback will inform the level before we finalize it.
The consultation lasts 33 days. After that, the discovery phase begins.
Disclosure: I am the CEO of InLinks and Waikay, both of which support the EntityMap standards proposal.
Additional resources:
Featured image: optimarc/Shutterstock
