What are data-* Attributes?

data-* attributes are custom HTML attributes used to store machine-readable metadata inside elements—supporting inline trust tagging, provenance tracking, and AI-ready publishing.

Full Definition

data-* attributes are part of standard HTML and allow publishers to attach custom, machine-readable values to any tag. While they don’t affect what users see, they can be read by scripts, crawlers, and AI systems to infer meaning, scope, or structure.

In the TrustPublishing framework, data-* attributes are used for:

Inline TrustTags for fact-level provenance
Attaching glossary term references to dynamic content
Flagging blocks for schema generation or Semantic Digest inclusion
Supporting custom logic for content rendering, scoring, or export

Why It Matters

Not every structured signal needs to live in JSON-LD or Microdata. Sometimes, you need to attach structure to a specific sentence, word, or block—without disrupting your layout or visible content.

data-* attributes let you:

Inject meaning anywhere in the DOM
Enable light-touch schema for dynamic systems
Flag elements for GPT-enhanced workflows or editorial pipelines

Examples

<span data-trust-id="cms-plan-2025">$0 Premium</span>

<div data-definedterm="Part D Deductible" data-source="CMS.gov">
   Learn how this deductible works.
</div>

<section data-faq="true" data-schema-type="FAQPage">...</section>

Use in Trust Publishing

data-* attributes are foundational to:

Applying TrustTags dynamically in CMS environments
Preparing block-level content for TrustDigest exports
Enabling semantic targeting for AI pipelines and training systems

They’re simple, lightweight, and render-proof—making them ideal for in-editor trust scaffolding.

In Speech

“data-* attributes are how we quietly tell the machine what’s important—without changing how it looks to a user.”

Related Terms

Microdata
TrustTags
Trust Publishing Markup Layer
Structured Signals