How Do Google Core Updates Affect AI Content Discovery and Citations?

What is a core update?

Google describes core updates as significant, broad changes to its search algorithms and systems, made several times a year and announced on its list of ranking updates. They do not target specific sites or pages. Google compares them to updating a list of favourite restaurants. New candidates appear, some move up, and those that move down are not necessarily bad; others have simply earned a place.

Most sites, Google says, do not need to worry about core updates and may not notice them. The ones that do notice are usually sites whose rankings on important queries moved sharply.

How do core updates reach Google's AI features?

• Retrieval uses core ranking systems. Google's guidance states that its AI features find relevant pages through its core ranking systems. When those systems are reassessed, the pool of pages available to AI Overviews and AI Mode changes with them.
• Citations overlap with rankings, but not completely. Ahrefs' 2026 analysis found about 38% of AI Overview citations came from the organic top ten, down from 76% in its July 2025 data. The rest came from lower-ranking pages, many far beyond the first page. A core update that moves a page from position 5 to 25 may reduce, but not end, its chances of citation.
• Related searches spread the effect. AI Overviews and AI Mode may search several subtopics for one question. A core update that affects your rankings for those subtopics affects your citations too, even if your main query holds steady.

How do core updates affect ChatGPT, Claude and Perplexity?

Indirectly. ChatGPT's search uses OpenAI's crawler and third-party search providers. Claude's web search draws on Brave's index. Perplexity runs its own index. None of them uses Google's rankings as its main index, but the lines are not clean: The Information reported in 2025 that OpenAI partly used Google results scraped through a third-party service for some ChatGPT answers, and a Peec AI study found 83% of the products in ChatGPT's shopping carousels come from Google Shopping results. Knock-on effects are possible:

• Traffic and links. A site that loses Google traffic may publish less, earn fewer links and lose visibility elsewhere over time.
• Shared signals. Other search engines assess similar qualities, such as helpfulness, originality and reputation. A site that a Google core update judges weaker may fare similarly elsewhere, but on its own schedule.
• Third-party sources. If the review sites or publishers that cite you lose rankings, answers that relied on them may change.

So a drop in ChatGPT or Perplexity citations in the same week as a Google core update is more often a coincidence than a consequence; check the other causes below before assuming either.

What else changes AI citations?

• Model updates. Ahrefs attributed part of the fall in top-ten overlap to AI Overviews moving to a newer Gemini model. Changes to the model behind an AI feature can shift citation patterns without any core update.
• Feature changes. New formats, such as AI local packs, citation displays or preferred-source options, change what is shown.
• Competitors' content. A competitor publishing a better answer can displace you in any engine.
• Your own changes: migrations, template changes, crawler blocks and content edits.
• Seasonality and demand: fewer searches mean fewer citations, not worse ones.

How do you tell whether a core update hit your AI visibility?

1. Check the dates. Note the start and end of the core update on Google's Search Status Dashboard.
2. Wait for it to finish. Google recommends waiting at least a full week after an update completes before analysing Search Console data.
3. Compare like with like. Compare a week after the update with a week before it started, for both organic rankings and AI Overview citations on your tracked questions.
4. Separate engines. If Google's AI citations fell while ChatGPT's and Perplexity's held, the core update is a likely cause. If all fell together, look for a site-wide change such as crawler access or a migration.
5. Separate query types. Drops concentrated in one topic or page type point to content quality in that area. Drops everywhere point to technical problems.
6. Check the search types. Google suggests analysing Web Search, Images, Video and News separately.

What should you do after a drop?

Google's advice distinguishes small and large drops:

• Small drops, such as from position 2 to 4, need no drastic action. Google recommends avoiding changes to content that is already performing well.
• Large, sustained drops, such as from position 4 to 29, call for a deeper assessment. Google points to its self-assessment questions on whether content is helpful, reliable and made for people rather than for rankings.

For AI citations specifically:

• Review the pages that lost citations. Do they answer the question directly, with specifics? Are they current?
• Look at the pages that replaced you. What do they offer that yours does not: data, examples, clearer structure, more recent facts?
• Improve substance, not signals. Rewrite for clarity and depth, add evidence, remove outdated sections. Cosmetic changes rarely help.
• Be patient. Google notes that recovery may take until the next core update, and possibly longer.

How engagement relates to AI visibility is covered in whether engagement metrics affect AI search visibility and citations. Whether publishing more often helps is covered in whether publishing frequency increases AI citations.

How do you reduce your exposure to core updates?

• Spread visibility across engines. Being cited in ChatGPT, Perplexity and Claude as well as Google means one engine's update hurts less.
• Build independent mentions, which help in every engine regardless of your own rankings.
• Keep content current and specific, the qualities core updates tend to reward over time.
• Track a fixed question set across engines, so you can see quickly where a change came from.

What might an illustrative case look like?

Consider an illustrative personal finance site that lost about a third of its AI Overview citations in the fortnight after a core update. Its ChatGPT and Perplexity citations were unchanged. Search Console showed large ranking drops for its credit card comparison pages but not its budgeting guides. The pages that replaced it in both results and AI Overviews had current rates tables, clearer eligibility criteria and named authors with credentials. The site rebuilt its comparison pages with monthly-updated tables and author bios, and did not touch the budgeting guides. It regained most of its citations over the following months, fully after the next core update.

What do people ask about core updates and AI?

Do core updates target AI-generated content?

Google says core updates are broad and do not target specific sites. Its guidance focuses on whether content is helpful and original, however it was produced.

Should we make big changes right after an update?

Not in the first week. Wait for the rollout to finish, analyse, then change what the evidence points to.

Can a core update improve our AI citations?

Yes. If the update favours your content, you may be retrieved and cited more often in Google's AI features.

Do other engines have core updates?

They change their systems too, but they do not announce them in the same way. Tracking across engines helps you spot their effects.