Skip to content

Your brand search now opens with an AI Overview citing Wikipedia, YouTube and LinkedIn. What to fix

AI Overviews now top most US branded searches, citing Wikipedia, YouTube and LinkedIn. What you control in each source, the opt-out trade-off and a brand audit.

By Published Updated 13 min read
Citations and accuracy, 13 min read — A large glowing glass panel in front of a smaller tile, with beams of blue-violet light converging on it from several directions.

The short answer

AI Overviews now appear on most US desktop branded searches, often in the top position, and cite Wikipedia, YouTube and LinkedIn more than other outside sources. Audit the overview for your brand queries, check every fact against each cited source, fix what you control on those platforms and your own site, and think hard before opting out.

Key takeaways

  • In Ahrefs' US desktop data, AI Overviews appeared on 82.91% of branded-keyword results on 29 September 2026, up from 61.3% across July.
  • In Ahrefs' panel of exact-name searches for 100 global brands, 63 showed an AI Overview on 30 September 2026, and 48 of those 63 (76.2%) sat in position one.
  • Of the 57 AI Overviews with visible sources in that 100-brand panel (July–September 2026), 71.9% cited Wikipedia, 38.6% YouTube and 22.8% LinkedIn.
  • You cannot write your own Wikipedia article, but you can make sure accurate, independent sources about you exist for editors to use.
  • The September Penske ruling shows a legal route will not fix a brand's overview soon; publishing and correcting sources will.
  • Opting out of Google's AI features removes your own pages as a source, not the overview built from everyone else's.

What Ahrefs found on branded searches

A search for your brand name now usually opens with an AI Overview, not your homepage. On 2 October 2026 Ahrefs published an analysis of 232.8 million US desktop Google result crawls from July to September 2026, split into branded and non-branded keywords.1 Ahrefs sells SEO software, including tools that track AI Overviews.

The rise was steep. As a monthly average, 61.3% of branded keywords triggered an AI Overview in July 2026 and 73.3% in September; on 29 September the share of branded results reached 82.91%.1 Non-branded keywords started higher, at 73.5% in July, and moved less.

Position matters as much as presence. Ahrefs also checked exact-name searches for 100 brands from Interbrand's Best Global Brands 2025 at 14 weekly checkpoints, a separate sample. By 30 September, 63 of the 100 showed an AI Overview, and 48 of those 63 (76.2%) sat in position one.1 For most of those searches, the first thing a person reads about a brand is a summary the brand did not write.

Branded searches, US desktop

The overview now comes before your link

82.91%of branded results with an AI Overview, 29 Sep 2026Ahrefs, Jul–Sep 2026 1
61.3%across July 2026, the start of the seriesAhrefs, Jul–Sep 2026 1
76.2%of AI Overviews in a 100-brand panel sat in position oneAhrefs, Jul–Sep 2026 1
Source: Ahrefs, AI Overviews on branded searches. The first two figures come from US desktop crawls of all branded keywords; the third from a panel of exact-name searches for 100 brands.

Why a branded overview matters more than it looks

Branded searches are made by people who already know your name. They are checking a price, looking for a login, reading reviews or deciding between you and a competitor. That makes them some of the most commercially valuable searches you have, and until this year you largely controlled the first result.

An AI Overview changes who frames the answer. It summarises what Google's systems found across several sources and puts that summary above your own link. If the summary says your pricing starts higher than it does, or describes a product you retired, the searcher reads that before reaching your site, and may not reach it at all.

Being the cited source also pays. Seer Interactive, a search agency, tracked 53 brands from January 2025 to February 2026. Brands cited in an AI Overview earned about 120% more organic clicks per impression than when they were not cited.6 That gap is a reason to make your own pages the ones the overview quotes.

Ahrefs quotes the SEO consultant Chris Long on the concern: competitors and external sources "can now control the narrative of your brand directly in the search results."1 Our view: for most companies the bigger risk is not a hostile competitor but stale facts. Old funding news, a former chief executive, a discontinued plan or a two-year-old review can all end up in the overview because they are what the cited sources still say.

The data has limits worth stating. The main series covers US desktop results only, so mobile and other countries may differ, and the source data comes from a 100-brand panel. It counts how often an overview appears, not how many clicks it takes away. Treat it as a measure of exposure, then check your own brand queries before deciding how much work to commit.

Which sources do branded overviews cite?

A small set of third-party platforms dominates. In Ahrefs' 100-brand panel (July–September 2026), 57 AI Overviews had a visible source list. Of those, 71.9% cited Wikipedia, 38.6% YouTube and 22.8% LinkedIn.1 Yahoo Finance and Facebook followed, at 15.8% and 10.5%.1

Outside domains cited in branded AI Overviews

Wikipedia71.9%
YouTube38.6%
LinkedIn22.8%
Yahoo Finance15.8%
Facebook10.5%
Share of the 57 AI Overviews with visible sources, in a panel of exact-name searches for 100 brands, citing each outside domain. Source: Ahrefs, Jul–Sep 2026.

Your own website still matters across AI engines more broadly. Profound, which sells AI visibility software, analysed 11.84 billion citations across eight engines from 16 April to 16 July 2026. About 57% went to company-owned websites, with Gemini at 69% and ChatGPT at 47%.4 Those are all citations, not only branded queries, so the two studies measure different things.

Read together, they suggest a practical split. On a branded Google search, outside reference platforms carry much of the description of who you are. Your own site supplies the specifics, such as products, prices and policies, when an engine looks for them. Both need to be accurate, and they need to agree with each other.

What you control on Wikipedia

You do not control your Wikipedia article, and you should not try to write it. Wikipedia's conflict-of-interest rules discourage people from editing articles about their own employer, and edits that read like marketing tend to be reverted. What you can influence is the material editors rely on.

An article is only as current as its independent sources. If the last substantial press coverage of your company is four years old, the article will describe the company of four years ago. Interviews, trade press, analyst notes and published filings give editors something citable. Your own press page counts for little on its own.

Where the article contains a factual error, use the talk page. Disclose who you are, state the error, and link an independent source that shows the correct fact. Then wait for an editor to act. It is slower than editing, and it works more often. Our guide to Wikipedia entries and AI answers covers notability, Wikidata and what to do if you have no article at all.

Our view: most brands should treat Wikipedia as a lagging mirror of their coverage. If the overview's Wikipedia-sourced sentences are wrong, the fix is usually upstream, in the independent coverage that does not yet exist.

What you control on YouTube, LinkedIn and your own site

These three are largely yours to manage, which makes them the fastest fixes on the list.

YouTube

YouTube citations in a branded overview usually come from your own channel, from reviewers or from explainer videos about your category. On your channel, make titles and descriptions state plain facts: what the product does, who it is for, what it costs if you publish prices. Upload accurate captions rather than relying on automatic ones, and add chapters so specific answers are easy to find.

Then look at the third-party videos that rank for your name. You cannot edit them, but you can comment with corrections, send reviewers current information and publish your own answer to the question they raise.

LinkedIn

LinkedIn shows up because company pages and posts are public, structured and frequently updated. Check that your company page's About section states what you do in the first two lines, with current headcount band, headquarters country and website. Leadership profiles should match the facts on your site. Our note on LinkedIn citations in B2B answers lists the fields that tend to be quoted.

Your own site

Make the facts an overview needs easy to lift. An About page that answers "what is this company" in the first paragraph, a pricing page with current prices in text, and a short facts page with founding year, ownership, locations served and leadership all help. Keep them dated. If AI helps draft them, check every fact by hand before publishing, as Google's updated guidance on AI content now asks. If you are named alongside a former name, a parent company or a confusingly similar brand, say how you differ.

Yahoo Finance and Facebook

The next two domains on the list reward basic housekeeping. Finance portals tend to appear for listed companies and draw on filings, company profiles and data feeds, so the fix is usually in the investor relations material those feeds read: a current business description, segment names and leadership. A Facebook page that has not been touched in years still carries an About section, opening hours and a description, and an overview may quote any of them. Update it or close it properly; a stale page is worse than none.

A tall pane of frosted glass in night blue, lit from behind, with several smaller glass tiles arranged in front of it, each catching a different band of blue-violet light.
An overview is assembled from several panes of light. Each source you fix changes what shows through.

Can a brand challenge Google over its AI Overview?

Not quickly, and not through antitrust law for now. On 1 October 2026 TheWrap reported that US District Judge Amit Mehta had dismissed Penske Media's antitrust claims over Google's AI summaries the day before, without prejudice. "An expectation is not an agreement," the judge wrote.2 Dismissal without prejudice means Penske could refile, so this is not the last word.

The case was about publishers' content being used in summaries, not about errors in a brand's description. Even so, it shows the limits of a legal route. A court case takes years, and the overview for your brand will be rewritten many times before one ends.

This is not legal advice, and a brand facing a defamatory statement should speak to a lawyer. For the ordinary case, where an overview repeats a stale or wrong fact, the effective route is the one Google's systems respond to: accurate sources, consistently stated, in the places the overview already reads. Our guide to correcting what ChatGPT gets wrong about your company applies the same logic to chat assistants.

Should you opt out of Google's AI features?

For most brands, no. Since 31 August 2026 Search Console's AI reports have been rolled out to sites worldwide, showing impressions from AI Overviews, AI Mode and generative AI features in Discover, without clicks. A property-level control that excludes a site's content and links from those three surfaces went worldwide the same day.3 We covered the launch in our note on Search Console's AI reports and the opt-out.

The trade-off on branded queries is uneven. The opt-out applies to your property. The overview for your brand name draws on Wikipedia, YouTube, LinkedIn and other sites too. Our view: opting out is more likely to remove your own accurate pages from the summary than to remove the summary, leaving third-party sources to describe you with less to correct them.

Opting out vs staying in, for branded searches

Opt out

  • Your pages stop feeding AI features
  • Third-party sources still describe you
  • Fewer ways to correct a wrong fact
  • Can suit publishers protecting content

Stay in and fix sources

  • Your facts page can be cited
  • Impressions visible in Search Console
  • Corrections on your site can flow through
  • Suits most brands on their own name
Illustrative. Our judgement for a typical brand; publishers with licensing concerns weigh it differently.

Publishers whose business is the content itself face a different calculation, and some will reasonably opt out. For a company whose goal is to be described accurately when someone searches its name, staying in and fixing the sources is usually the better position.

A branded-results audit you can run this month

An audit of your branded results takes a few days and turns a vague worry into a list of owners and fixes. The method is the same whether you run it by hand or with a tool.

Start with a list of branded queries, not just the brand name. Add the brand with "pricing", "reviews", "alternatives", "vs" and your main competitor, "careers", "login", "is it legit" and each main product name. Capture the overview for each, record every cited source, and mark each factual statement as correct, outdated or wrong. Then give each source an owner.

The branded-results audit

  1. List the queries

    Brand name plus pricing, reviews, alternatives, competitor, careers and products.

  2. Capture the overviews

    Record the text and every cited source, with the date and device.

  3. Mark each fact

    Correct, outdated or wrong, checked against your own records.

  4. Assign owners

    Wikipedia via coverage, YouTube and LinkedIn by your team, your site by web.

  5. Re-check monthly

    Same queries, same method, compared with the last run.

Illustrative. Our method; adapt the query list to your brand.

Prioritise by intent. Queries with "pricing", "reviews" and "vs" sit closest to a purchase, so a wrong fact there costs the most. "Careers" and "is it legit" queries shape how candidates and cautious buyers see you. Fix the facts behind the high-intent queries first, then work outwards.

Run each capture more than once. Overviews change between days and between sessions, and one capture can mislead. When we run this as part of an audit, our AI answer monitor covers six engines (ChatGPT, Gemini, Perplexity, Claude, Copilot and Google AI) and runs each prompt on three separate days. It uses official APIs where engines offer them and a licensed AI-answer data provider otherwise. Our accuracy audit guide lists the facts buyers check most often.

Beyond the five domains: why mentions matter

The cited domains are the visible part. Behind them sits the wider body of coverage that tells search systems what your brand is. In an earlier Ahrefs study of 75,000 brands, published in May 2025 and now more than a year old, branded web mentions correlated with AI Overview visibility at 0.664, against 0.218 for backlinks.5 That is a correlation, not proof that mentions cause visibility, and the data is dated.

It still fits what the branded data shows. The sources an overview cites are the ones that describe you most often and most recently. A brand that is discussed accurately in trade press, on review sites and in industry directories gives every one of those sources better material. Our piece on review sites and directories covers the platforms that matter by industry.

Our view: treat branded overviews as a public summary of your coverage. If the summary is thin or wrong, the most durable fix is more accurate, independent material about you, published where these systems already look.

What to do this week

Search your brand name and five branded variations, on desktop and on a phone, and save what the overview says. Check each statement against your own records. Note which sources were cited.

Then fix the fastest things first: your About, pricing and facts pages, your LinkedIn company page and your YouTube descriptions. Start a list of independent coverage that would give Wikipedia editors current material. Open Search Console and look at the AI features report for your brand queries. Leave the opt-out alone until you have read what it covers and decided what you would lose.

Report the result to leadership as a short table: each branded query, whether an overview appeared, which sources it cited, which facts were wrong and who owns each fix. Repeat it next month with the same queries. A brand's search result used to be a list of links it could mostly predict. It is now a summary that changes as its sources change, and the only reliable way to know what it says is to look, on a schedule, and keep the record.

Sources

  1. Ahrefs, AI Overviews on branded searches: US desktop, Jul–Sep 2026; source shares and position from a separate panel of 100 brand-name queries, 14 weekly checkpoints (Oct 2026)
  2. TheWrap, Penske Media lawsuit over Google AI search summaries dismissed (1 Oct 2026)
  3. Search Engine Journal, Search Console AI reports rolled out worldwide (31 Aug 2026)
  4. Profound, where AI citations come from: 11.84B citations, 8 engines, 16 Apr–16 Jul 2026 (Jul 2026)
  5. Ahrefs, brand mentions and AI Overview visibility, correlational (May 2025)(dated)
  6. Seer Interactive, AI Overviews and CTR, 2026 update: 53 brands, Jan 2025–Feb 2026 (Apr 2026)

Questions readers ask

  • In Ahrefs' analysis of US desktop results from July to September 2026, AI Overviews appeared on 61.3% of branded-keyword results across July and 82.91% on 29 September. The data covers US desktop only, so check your own brand queries on mobile and in other countries before drawing conclusions about your exposure.

Keep reading

Free, in two minutes. Enter your domain and we'll score it against three competitors across six engines.

No account needed. The report is emailed within 24 hours.