New reporting from Search Engine Journal, based on Ahrefs data, suggests that ChatGPT’s cited sources in the US are increasingly tilted toward primary-source websites such as government, research and regulatory domains, while secondary explainer content is becoming less prominent in citations.

Although the data referenced in the report is focused on US citation patterns, the development is relevant for European businesses that rely on organic visibility, thought-leadership content and AI-assisted discovery. If AI systems increasingly surface original documents, datasets and official guidance, companies may need to rethink how they create content for both human readers and AI search environments.

What happened

According to Search Engine Journal’s coverage of Ahrefs data, ChatGPT’s most-cited websites in the US have shifted toward primary sources. The report says government health, research and regulator sites gained visibility in citations, while more interpretive or explanatory content lost ground.

The article points to a broader change in how AI systems may select references: instead of relying as heavily on summarised third-party articles, they may be leaning more often on original source material when producing answers.

That does not necessarily mean explainers have become unimportant. Secondary content still plays a role in helping users understand technical, legal or industry-specific topics. But if citation behaviour is moving toward official or original documents, businesses may need stronger source alignment in their content strategies.

Why it matters for European businesses

For European companies, especially those investing in SEO, content marketing and AI visibility, this trend could affect how content performs in generative search and AI assistants.

• Authority may increasingly depend on source quality: Content that clearly references official guidance, original research, product documentation or regulatory text may have a stronger chance of being reflected in AI-generated answers.
• Traditional explainer content may face more pressure: Businesses that publish high volumes of derivative blog content without adding original insight, data or evidence may find it harder to stand out in AI-mediated discovery.
• Regulated sectors could see bigger effects: In fields such as healthcare, finance, privacy, cybersecurity or compliance, AI tools may prefer official sources because accuracy and traceability matter more.
• Brand visibility may shift: If AI systems cite source material instead of intermediary publishers, traffic and brand exposure patterns could change for companies that depend on informational search visits.

This is particularly important as businesses across Europe adapt to search journeys that no longer begin and end with a standard list of blue links. AI assistants can summarise, compare and cite information directly, which changes the role of corporate websites, blogs and knowledge hubs.

Who may be affected

• Marketing teams and SEO managers building content strategies around organic discovery and topical authority.
• E-commerce businesses that use educational content to support category discovery, product research and conversion.
• B2B companies publishing industry explainers, compliance guides or solution comparison pages.
• Digital agencies and publishers whose content models rely on ranking and visibility for informational queries.
• IT, legal and compliance-focused organisations that produce documentation, policy pages or technical guidance.

What companies should consider

• Review content quality at the source level: Check whether articles cite authoritative documents, original research, standards, public datasets or first-party expertise.
• Publish more original material where possible: Case studies, benchmark data, technical documentation, methodology pages and expert commentary may become more valuable than generic summaries.
• Strengthen content traceability: Make it easier for both users and AI systems to identify who wrote content, what evidence supports it and when it was updated.
• Align explainers with primary references: Businesses do not need to stop publishing educational content, but they may need to connect it more clearly to source documents and verifiable claims.
• Monitor AI visibility separately from traditional rankings: Standard SEO reporting may not fully show whether a brand is being surfaced, cited or bypassed in AI-generated answers.

The reported shift does not by itself establish a universal rule for all markets or all AI systems, and the source article refers specifically to US-focused citation data. Still, for European organisations preparing for AI-influenced search, the message is practical: content that is closer to original evidence and clearer in its sourcing may be better positioned than purely derivative explainers.