What is Structured Data / Schema Markup?
Structured data (schema markup) is machine-readable code — typically JSON-LD following the schema.org vocabulary — added to a page to explicitly state facts like what an entity is, who authored a piece of content, or how a page is organized, rather than leaving a model to infer them from prose.
Common types include Organization and WebSite (entity identity), Article (authorship and publish/update dates), FAQPage (question/answer pairs), BreadcrumbList (site hierarchy), and Person (author credentials). Search engines have used structured data for rich results for years; the same markup serves AI systems for a related but distinct reason — it states facts plainly instead of requiring a model to parse them out of unstructured text.
Structured data should only describe what's genuinely present and visible on the page. Markup that states facts a page doesn't actually contain is a validation risk and, more importantly, can read as an attempt to manipulate rather than clarify.