Duplicate Content
Duplicate content is text that appears, word for word or near enough, at more than one address — on the same domain or across different ones. A search engine or an AI system then has to pick a single version to trust, and it has no reliable way to know which copy is the source. The practical effect is that your own signals get split across the copies instead of adding up behind one page.
Why it matters
When the same passage lives at several URLs, ranking systems and citation systems both face the same problem: which one do we point to? Links, mentions and click history that should all reinforce one page instead get divided among two, three or more near-identical ones, so none of them performs as well as a single consolidated page would. For AI answer engines the cost is sharper still. An Ahrefs study on ChatGPT citations found that only 6–8% of the URLs it cites overlap with Google's top 10 for the same query, and roughly 80% of ChatGPT-cited URLs don't rank in Google's top 100 at all — meaning AI systems are running their own source selection, not just reading rank position. A system doing its own selecting needs an unambiguous, attributable source even more than a search engine does. Duplication removes exactly that: if three pages say the same thing, none of them is clearly "the" source, and the assistant has no reason to pick yours.
How it arises
Almost none of it is intentional. Technical duplication is the most common cause: the same page reachable with and without "www", with and without a trailing slash, under both HTTP and HTTPS, or through filter and sort parameters that generate a new URL for every combination in a shop. Print-friendly versions, session IDs appended to a URL, and staging copies left crawlable add more. Content-level duplication comes from a different habit: pasting in a manufacturer's product description unchanged, syndicating one article across several of your own pages, or reusing one boilerplate paragraph across dozens of location or service pages. In every case the pattern is the same — a system finds identical substance at multiple addresses with no signal telling it which one is original.
How you avoid it
The standard fix is the canonical tag: a line in a page's source declaring which URL is the authoritative version, so that any signals earned by the copies get attributed to that one address instead. Permanent 301 redirects handle the cases where a duplicate URL shouldn't exist at all — point it straight at the real page. Consistent internal linking, always using the same URL form, keeps you from creating new duplicates by accident. None of this, notably, requires special AI-facing markup: Google's own guidance on optimizing for AI Overviews and AI Mode states plainly that no special schema, structured data, or AI-specific file is required, and explicitly advises against writing separate content "for AI." The fix for duplicate content is the same one that has always applied — canonicalize, redirect, and write your own words instead of reusing someone else's.
Relation to citability
An AI assistant naming a source is making a judgment about which page most clearly and reliably answers a question. Duplicate content works against that judgment by making your own best page interchangeable with its own copies. Consolidating duplicates doesn't manufacture citations on its own, but it removes a self-inflicted obstacle: a single, canonical, well-written page gives an AI system one clear thing to point to, rather than three ambiguous options split across your own site or a partner's. That's a precondition for citability, not a substitute for having something worth citing in the first place.
Example
A kitchenware retailer lists the same espresso machine in its Espresso, Small Appliances and Deals categories. Each category page gets its own URL, and all three carry the identical description copied from the manufacturer's data sheet. A search engine now has three near-identical pages to choose from and ranks none of them particularly well; an AI assistant asked to recommend an espresso machine skips all three in favor of a review site with its own tested, original write-up. Once the retailer sets a canonical tag pointing to the main product page and replaces the copied text with its own description, the ranking signals consolidate on one URL and that page becomes something an AI system can actually point to.
Common questions
Does Google penalize duplicate content?
Not with a manual penalty, as a rule. Google filters duplicates and shows one version, and the others are simply set aside — the cost is diluted visibility, not a punishment. Sanctions only come into play with content copied at scale specifically to manipulate rankings.
Is republishing my own article on a partner site duplicate content?
Yes — that's cross-domain duplication. Either have the partner site add a canonical tag pointing back to your original, or publish a substantially rewritten version there so your page stays the clear source.