Using ld+json for Structured Data in Generative AI
Using ld+json is essential for generative AI to understand and reference your content. Learn how to structure your data.
With the rise of Generative Artificial Intelligence, the way information is discovered, interpreted and presented is undergoing a profound transformation. Tools such as ChatGPT, Gemini and Copilot are no longer limited to answering based on traditionally indexed data. They interpret, connect and present content based on its semantic structure, relational context and source authority.
In this new landscape, using structured data with ld+json based on Schema.org is not just an SEO best practice, but a strategic requirement for your content to be understood and cited by generative AI agents.
What is ld+json?
ld+json (Linked Data in JSON) is a lightweight format for describing structured data directly in a page's HTML, without affecting its visual rendering. This markup helps search engines and AI systems understand a page's content as entities and relationships, rather than just blocks of text.
Why is structured data crucial for generative AI?
Generative AI works through language models that don't just “read,” but interpret the deeper meaning of published information. They need to identify:
- Who is speaking (authority/authorship);
- What it's about (semantic context);
- How it connects to other information (ontological relationships);
- Whether it's trustworthy, up to date and original (EEAT signals: Experience, Expertise, Authoritativeness, Trustworthiness).
Without structured data, AI has to infer this from context, increasing the chances that your content will be misinterpreted, excluded or overlooked in generated answers.
Direct benefits of using ld+json for AI and modern search engines
1. Makes it easier for generative AI to extract and cite your content
Models such as ChatGPT or Bard, when connected to the web, prefer sources that make their metadata explicit (author, content type, date, category, etc.). This increases the likelihood that your content will be used as a trusted reference in responses.
2. Improves rankings on search engines that already integrate AI
Google, Bing and others already use AI to display more accurate and personalized results. Pages with well-implemented schema tend to stand out more in rich results and featured snippets.
3. Builds trust and authenticity
Marking up Person, Organization and Article helps AI understand who is publishing the content, the reputation of that entity and the context in which it belongs.
4. Enables contextualized answers
AI can interpret data such as location, dates, related topics, user intent and possible actions, providing more refined answers with specific, personalized context.
Most relevant structures for blogs, portals and corporate websites
The most commonly used structures, and those recommended for generative AI, include:
| Schema type | Purpose |
|---|---|
WebSite | Defines the site's global entity |
Organization | Identifies the company or brand responsible |
Person | Content author, with name, image and profile |
Article | Editorial content structure (blog post, news article, etc.) |
WebPage | Information specific to the page being accessed |
BreadcrumbList | Hierarchical path to the content |
FAQPage | Questions and answers for use in snippets/AI |
HowTo | Step-by-step instructions for completing a task |
Event | Complete information about an in-person or online event |
Example of ld+json for an article
{
"@context": "https://schema.org",
"@type": "Article",
"headline": "The Importance of Using ld+json for Structured Data in an Era Dominated by Generative AI",
"author": {
"@type": "Person",
"name": "Rodrigo Neves",
"url": "https://vitamina.digital/author/rodrigo-neves/"
},
"publisher": {
"@type": "Organization",
"name": "VitaminaWeb Digital Production Company",
"logo": {
"@type": "ImageObject",
"url": "https://vitamina.digital/logo.png"
}
},
"datePublished": "2025-04-20",
"description": "Learn how using structured data in ld+json can increase your content's visibility to generative AI and advanced search engines."
}
Strategic considerations
- Don't rely only on visual presentation or traditional SEO. The AI tools powering new answer engines depend on logical and semantic structure.
- Update your content and schemas regularly. Outdated data reduces trustworthiness and rankings.
- Test your pages with the Rich Results Test and the Schema Markup Validator.
The shift from traditional SEO to AI-driven semantic SEO is underway, and using ld+json with structured data is one of the pillars of this evolution. Companies and professionals who want their content recognized, referenced and ranked by generative AI need to adopt this practice as part of their digital strategy.
It's no longer just about being found on Google. It's now about being interpreted and cited by artificial intelligence that is shaping how people consume content.
Read also: