Press releases have outlived a dozen predicted deaths, so it is fair to ask whether they still do anything useful now that AI Overviews and chat assistants sit between your brand and the searcher. The instinct among marketers splits two ways: some treat a wire blast as a reliable visibility lever, others write PR off entirely. Both are wrong, because they judge the press release itself. The thing that matters is what the release earns. Genuine, authoritative coverage creates the mentions AI systems rely on; a syndicated boilerplate blast creates noise. Get that distinction right and the question answers itself.
The short answer
It depends on the pickup, not the release. AI Overviews and chat assistants lean heavily on brand mentions across credible sources, and Ahrefs, studying 75,000 brands, found that web mentions correlate with AI visibility far more strongly than backlinks do (Ahrefs). A press release that earns genuine, authoritative coverage generates those mentions and helps your AI presence. A wire blast that just duplicates the same boilerplate across low-value sites generates low-trust copies that do little. So press releases can boost AI presence, but only through the quality of coverage they earn.
Why this question matters now
The stakes have changed. In classic SEO, a press release was often chased for links, and even a syndicated blast might drop some links that nudged rankings. In AI search, that logic weakens, because links are not the currency, a shift we cover in do ChatGPT links count as backlinks for SEO. What AI systems reward is being credibly talked about, which reframes PR entirely: the goal is not to place a link but to earn a mention in a source the model trusts. That is a higher bar than a wire drop, and it is the bar that matters now.
How AI Overviews decide what to surface
To see why pickup beats the release, look at how AI systems choose what to show. They draw on a vast reading of the web and favour sources that credibly and repeatedly discuss a topic. Ahrefs’ analysis of the most-cited domains in ChatGPT shows the model leaning on authoritative and widely referenced sources across millions of queries (Ahrefs), and Google notes that appearing in AI features draws on your normal web presence (Google Search Central). The common thread is credibility and repetition across trusted sources, which is precisely what earned coverage provides and what boilerplate syndication does not.
The key distinction: the release versus the pickup
Here is the whole post in one idea. A press release is an input; the coverage it earns is the output; and only the output moves your AI presence. When a journalist reads your news and writes a genuine story, or an authoritative outlet covers it in its own words, that creates an original, trusted mention. When a wire service copies your exact text onto two hundred low-value syndication pages, that creates two hundred near-identical, low-trust copies. The first is the mention currency AI rewards; the second is largely ignored. Same release, completely different value, decided entirely by what it earned.
Why a wire blast alone does little
Duplicate boilerplate is the problem. Models and search engines have long been good at recognising syndicated, identical content and discounting it, so a release republished verbatim across many thin sites does not read as many independent endorsements. It reads as one press release, copied. There is no independent corroboration in a hundred copies of the same words on sites nobody trusts. So the raw syndication count that wire services sell as reach is close to a vanity metric for AI visibility purposes; it is volume without the trust and originality that actually count.
There is a subtler failure too. Because the syndicated copies are identical, they give the model nothing new to learn about you. A mention only teaches the model something when it adds context, an angle, a use case, a verdict, and boilerplate you wrote about yourself adds none of that, no matter how many times it is copied. It is your own marketing language echoed back, which is exactly the self-referential signal models are least persuaded by. A hundred copies of your own words are still just your own words, and the model already knows what you say about yourself. What it wants is what others say.
Why earned pickup does a lot
Genuine coverage is the opposite. When real outlets cover your news in their own words, you get original, varied, credible mentions across sources the models weight highly. Those are exactly the signals Ahrefs found correlate with AI visibility, mentions across diverse, trusted sources rather than links or raw volume (Ahrefs). One authoritative pickup can be worth more than a thousand syndicated copies, because it is a real, independent voice associating your brand with the topic. This is the same corroboration logic that makes reviews matter, explored in do G2 reviews affect ChatGPT software rankings.
Earned coverage also tends to carry the specific, contextual language that helps a model understand what you are for. A journalist explaining why your product matters, who it suits, and how it compares gives the model exactly the kind of nuanced, third-party framing it can draw on when answering a specific question. That framing is something you cannot manufacture by repeating your own tagline; it has to be earned from someone with no incentive to parrot your marketing. This is why a single thoughtful feature in a trusted outlet can shape how AI describes you for months, while a syndication blast leaves no such fingerprint.
What the data says
The evidence points consistently in one direction. Web mentions, not backlinks, are the strong correlate of AI visibility across 75,000 brands (Ahrefs), and the most-cited sources skew toward authoritative and widely referenced domains (Ahrefs). There is also an indirect link benefit worth noting: earned coverage that includes real links can lift your Google rankings, and a large share of AI Overview citations still come from pages ranking in the top 10 (Ahrefs). So quality PR helps twice, directly through mentions and indirectly through rankings, while a boilerplate blast helps through neither.
PR that works versus PR that does not
This table separates the two clearly.
| Attribute | PR that helps AI presence | PR that does not |
|---|---|---|
| Output | Original, earned coverage | Duplicate syndicated copies |
| Source quality | Authoritative, trusted outlets | Low-value syndication sites |
| Wording | Journalists write in their own words | Verbatim boilerplate everywhere |
| What it creates | Credible, independent mentions | One release, copied many times |
| AI value | High, real corroboration | Low, discounted as duplicate |
The indirect ranking benefit
Do not overlook the second path. Because AI answers, especially AI Overviews, still lean on pages that rank well, PR that earns genuine coverage and links can improve your rankings, which in turn raises your odds of being cited (Ahrefs). This is the same mechanism by which classic authority feeds AI visibility: the coverage lifts your ranking, and the ranking feeds the citation. It is another reason to chase real pickup, which can earn meaningful links, over syndication, which rarely does.
The two paths also reinforce each other over time. A genuine feature in a respected outlet earns a mention today, may earn a link that lifts your ranking, and often gets referenced by other writers later, compounding into more mentions. A syndicated blast has no such second life; the copies sit inert and are never cited by anyone. So the gap between earned and syndicated PR widens the longer you look, which is worth remembering when a wire package promises impressive-sounding distribution numbers for a low price. Cheap reach that compounds to nothing is more expensive than it looks.
How to run PR that actually helps
Aim the whole effort at earning genuine coverage. Have real news worth covering, not a manufactured announcement, because journalists and their readers can tell. Target the specific, authoritative outlets your audience and the models trust, rather than maximising raw distribution. Make it easy for a journalist to write an original story by giving them substance, data, and a real angle. And measure success by the quality of coverage earned, not by syndication counts. The broader discipline of earning citations, of which PR is one lever, is laid out in how to get cited in ChatGPT.
What not to do
The losing move is to equate a press release with a result: paying for maximum wire distribution and counting the syndicated copies as impact. That optimises the vanity metric and ignores the trust metric. Equally, do not manufacture fake news or spin thin announcements into releases, which earns no genuine coverage and can erode credibility. And do not treat PR as a link-buying scheme, since links are not the AI currency anyway. Chase the real story and the real pickup, not the distribution count.
A worked example
Say a company puts out a funding announcement two ways. Version A goes out as a wire blast and lands, verbatim, on 150 syndication sites nobody reads. Version B is pitched to three respected industry outlets, two of which write original stories about the raise. Months later, when someone asks an AI about companies in that space, the model has three things: 150 identical low-trust copies it discounts, and two genuine, authoritative articles it trusts. The two earned pieces are what make the company more likely to be mentioned; the 150 copies contribute almost nothing. Same news, and the difference in AI presence comes entirely from earning real coverage rather than buying reach.
Common misconceptions
The biggest misconception is that issuing a press release, by itself, boosts AI presence; it is the earned coverage that does. The second is that syndication reach equals impact, when duplicate copies are discounted and only original mentions count. The third is treating PR as link building, when mentions, not links, are the AI currency. The fourth is assuming any coverage helps equally, when authority and originality of the source are what matter, a selectivity related to why your brand is missing from AI recommendations.
The bottom line
Do press releases still boost brand presence inside AI Overviews? Yes, but only through the genuine, authoritative coverage they earn, not through the release or its raw syndication. AI systems reward credible mentions across trusted sources, and the data shows mentions beat links and volume. So run PR to earn real stories in outlets the models trust, measure by the quality of pickup rather than distribution counts, and enjoy the double benefit of direct mentions and the rankings that earned coverage can lift. Judge the pickup, not the press release, and PR earns its keep in AI search.
Frequently asked questions
Do press releases directly get me into AI Overviews?
Not directly. The release itself is not what AI systems reward. What helps is the genuine, authoritative coverage a release can earn, because that coverage creates the brand mentions across trusted sources that AI answers lean on. Judge PR by the pickup, not the press release.
Is a wire-service press release blast worth it for AI visibility?
On its own, usually not much. A wire blast tends to create many duplicate copies of the same boilerplate on low-value sites, which are low-trust and add little. It can seed awareness, but the value comes only if it leads to real, original coverage in credible outlets.
Why do mentions matter more than the release itself?
Because AI systems are drawing on how often and how credibly your brand is discussed across the web. Ahrefs found mentions correlate with AI visibility far more than backlinks. A release matters only insofar as it generates genuine mentions in sources the models trust.
What kind of PR actually helps AI presence?
PR that earns real, original coverage: a journalist writing a genuine story, an authoritative outlet covering your news in its own words, an expert quoting you. That creates trusted, non-duplicate mentions. Aim the release at earning that, not at maximising raw syndication counts.
