AI search glossary
What is Wikidata? Items, properties and why AI reads it
Wikidata is a free, collaborative, multilingual knowledge base run under the Wikimedia umbrella that stores structured facts about entities as items and statements, released under a CC0 public domain dedication so anyone can reuse them.
Also called: Wikidata item, Q-ID, wikidata.org
Why it matters for AI search
Because Wikidata is open, structured and free to reuse, it is one of the most widely copied sources of entity facts on the web. Google announced in 2014 that it would close Freebase in favour of Wikidata, its Knowledge Graph API docs point to Wikidata dumps, and voice assistants have used its data. An accurate, referenced item helps search engines and AI models identify an entity correctly.
How Wikidata is structured
Wikidata stores facts, not articles. Its building blocks:
| Part | What it is | Example |
|---|---|---|
| Item | An entity, with an ID starting with Q, a label, a description and aliases | Q42, Douglas Adams |
| Property | A kind of relationship, with an ID starting with P | P69, educated at |
| Statement | An item, a property and a value | Q42, P69, St John's College |
| Reference | The source behind a statement | A university record or news article |
| Sitelink | A link to the matching page on Wikipedia or another Wikimedia site | English Wikipedia article |
Properties that point to other databases are called identifiers: a company's official website, its LinkedIn page, its stock ticker. These are what tie one entity together across the web. Anyone can edit, and registration is optional. As of October 2026 Wikipedia reports more than 123 million data entities in Wikidata.
Who uses it
Wikidata supplies structured data to Wikipedia and other Wikimedia projects, and to anyone else. When Google announced in December 2014 that it would close Freebase, it moved that data toward Wikidata, and its Knowledge Graph Search API documentation still recommends Wikidata dumps for graph data. Voice assistants such as Siri and Alexa have used it. Because it is CC0, it is easy for any search engine, data company or AI developer to ingest.
Getting a business onto Wikidata
- Check notability first. You need a Wikimedia sitelink, serious public references that clearly identify the entity, or a structural need. Self-promotion without references tends to be removed.
- Add facts with references. Instance of (business), official website, inception date, headquarters location, industry, founders. Cite independent sources, not only your own site.
- Add identifiers for profiles you control.
- Link back. Add the Wikidata URL to the
sameAslist in your Organization schema.
Do not invent facts or add marketing language. Editors patrol changes, and wrong data spreads to every system that copies it, which is one source of AI hallucinations about brands.
Related ideas
Wikidata is an open knowledge graph. It is one of the sources behind a Google knowledge panel, and keeping it accurate is part of entity SEO.
Questions
- Who runs Wikidata?
- Wikidata is hosted by the Wikimedia Foundation, the non-profit behind Wikipedia, and mainly developed by Wikimedia Deutschland. It launched on October 29, 2012.
- Can I create a Wikidata item for my company?
- Only if it meets Wikidata's notability policy: the item has a valid sitelink to a Wikimedia page such as Wikipedia, or it refers to a clearly identifiable entity that can be described with serious, publicly available references, or it fills a structural need. Items without references are often deleted.
- Is Wikidata data free to use?
- Yes. Structured data in Wikidata is released under the Creative Commons CC0 public domain dedication, so it can be reused, including commercially, without asking permission.
- What is a Q number in Wikidata?
- Each item has an identifier starting with Q. Douglas Adams, for example, is Q42. Properties start with P, such as P69 for educated at.