Duplicate Content in the Age of AI

Anyone who wants to be found on Google has long focused on unique, high-quality content. But with the growing use of ChatGPT, Google AI and other AI systems, the topic is taking on a new dimension: it's no longer just about being found – it's about being recognized and cited by an AI as a source.

A recent study by Shero Commerce shows just how big this problem already is. It examined product information from 1,000 Shopify stores and analyzed which sources Google, ChatGPT and Perplexity use for typical user queries. The findings can generally be applied to other websites and content management systems such as TYPO3 as well.

When the AI Knows Your Content – But Cites Someone Else

One of the study's most interesting findings: an AI can be well aware of a brand and recommend it in its answer without ever naming the company's own website as the source.

This creates a new problem for companies: their own information may well make it into an AI answer, but the user is then referred to a different website.

Duplicate Content Makes Source Attribution Harder

One possible reason for this is duplicate content.

The problem from an AI's perspective: if the same text exists on numerous websites, there's little reason to treat the original provider's website as the authoritative source.

A magazine, trade portal, or other website that instead publishes an original, genuinely helpful text on the same topic can be the more attractive, citable source for an AI system.

Classic SEO and AI Visibility Are Not the Same Thing

One distinction matters here: the study does not prove that duplicate content leads to a Google penalty or directly causes worse rankings. Rather, it examined how often content appears in duplicate form and which websites AI systems subsequently use as sources.

This also changes the underlying question in search engine optimization.

This is exactly where GEO (Generative Engine Optimization) and AEO (Answer Engine Optimization) come in.

Not the Most Content – the Most Original Content

The study also shows that sheer volume of text is not the deciding factor.

Websites that were actually used as sources stood out instead for containing specific, original information. Typical traits were individual descriptions, concrete answers to users' questions, and content that was directly present in the page's HTML.

That's an important insight for the future content strategy of any website.

Content Must Be Technically Accessible to AI Systems

Beyond content quality, technical implementation also plays a role.

The study specifically looks at the HTML content that crawlers can actually retrieve. Content that is generated exclusively via JavaScript after the fact can be significantly harder – or entirely impossible – for certain AI crawlers to access.

For an AI-friendly website, important information should therefore not only be visible to human visitors, but also technically well-structured and machine-readable.

What Does This Mean for TYPO3 Websites?

This topic is by no means limited to online shops. The study's authors themselves point out that the underlying problems are not Shopify-specific.

Especially on extensive TYPO3 websites, information is often found in multiple places:

Companies and organizations should therefore make sure their own website is the most comprehensive and highest-quality original source on their topics.

Instead of generic marketing copy, concrete content becomes more important: expert information, proprietary data, definitions, how-tos, FAQs, study results, experience reports, or detailed answers to typical user questions.

TYPO3 Provides a Good Foundation for This

TYPO3 makes it possible to build extensive information architectures in which individual topics are covered on their own, clearly structured subpages.

From a technical standpoint, this should include, among other things:

The combination of classic SEO and GEO/AEO becomes especially interesting here. A technically clean website remains important – but at the same time, its content must be crafted so that search engines and AI systems can understand what information a page offers, and why this particular website is a relevant original source.

Conclusion: From "Being Found" to "Being Considered a Source"

The growing spread of AI-generated answers is changing the requirements for websites.

It's becoming less and less sufficient for information to simply exist somewhere on a company's own website. Businesses should publish content that is original, professionally relevant, technically accessible, and clearly attributable to their site.

Duplicate content therefore takes on an additional significance – not necessarily as a classic SEO problem, but as a problem of source attribution.