How to Use Press Coverage to Increase Your Brand’s LLM Presence

Press coverage helps a brand’s presence in AI answers when it is independent and corroborated, and much less when it is merely plentiful. A 2026 experiment by Authoritas makes the point sharply. It tested 11 fake expert personas that had been featured across more than 600 UK media articles before being exposed. When nine AI models were asked 55 topic questions, such as who the UK’s leading experts on home security are, none of the fake personas appeared. Volume of coverage alone did not make them part of the answer. When the models were asked about the fake names directly, though, they treated them as credible authorities 29.5% of the time, sometimes inventing career histories. Coverage shapes what models say about a name once asked, even when it does not earn a place in answers about a topic.

That combination defines the job: earn coverage that establishes real standing, and make sure what the coverage says is accurate.

What press coverage can do

It corroborates you independently. A company’s own website is a self-description. Reporting by a publication that chose to cover the company is outside evidence. Wikipedia’s notability guideline builds on exactly this distinction: a topic is presumed suitable for an article when it “has received significant coverage in reliable sources that are independent of the subject.” Coverage that meets that standard can lead to a Wikipedia article, which in turn is a source models have learned from and that AI summaries cite.

It enters the public record models learn from. Training data is drawn from large amounts of public text and, at least in the training process OpenAI described for GPT-3, weighted toward sources judged higher quality. The GPT-3 paper says datasets “we view as higher-quality are sampled more frequently.” Reporting in reputable publications is part of the web a model may learn from. How much weight any given publication carries in today’s models is not public.

It can be retrieved at answer time. When an AI system searches the web, news articles are among the pages it can find and cite, alongside your own site.

What press coverage cannot do

Manufactured mentions do not help. Google’s AI optimization guide says inauthentic “mentions” are not a way to improve visibility in its AI features. Paid placements dressed as editorial, mass-distributed press releases and articles placed purely to create mentions do not provide the independent corroboration that matters.

Volume does not substitute for standing. The fake personas in the Authoritas experiment had hundreds of articles and still did not appear in topic answers. Real experts, the study notes, built “deep, corroborated brand entities,” with consistent profiles and third-party validation.

Coverage can be misrepresented. AI tools do not always attribute news accurately. The Tow Center for Digital Journalism tested eight AI search tools in early 2025 by asking them to identify the source of news excerpts, and found they answered more than 60% of queries incorrectly. A study coordinated by the European Broadcasting Union and the BBC, published in October 2025, found at least one significant issue in 45% of more than 3,000 AI assistant responses about news. Being covered is not the same as being described correctly.

How to earn coverage that helps

1. Give reporters something worth covering. Original data, a clear expert view on a live question, a notable result or a useful tool. Coverage earned on substance is more likely to be specific, and specific coverage is easier for a model to describe correctly.
2. Be precise about the facts you share. Provide your company name, the spelling of key people’s names, what you do and where, in the same form every time. Inconsistent details across articles can create conflicting records.
3. Seek reputable, independent outlets relevant to your field. Coverage in sources that a model’s makers or Wikipedia editors would treat as reliable is likely to count for more than a large number of low-quality placements.
4. Pair coverage with expertise. If your people are quoted as experts, make sure their own profiles, author pages and publications support what the coverage says about them.

How to keep coverage accurate

• Correct errors at the source. If an article gets a key fact wrong, ask the publication for a correction. A corrected article is better material for both training and retrieval than a wrong one.
• Keep your own pages consistent with your coverage. Your About page, leadership bios and press page should state the same facts the coverage reports.
• Check what AI tools say about you. Ask several engines direct questions about your company and its people, repeat them over time, and note any errors. The Authoritas results show that models can repeat what they are told about a name, so wrong information in circulation is worth tracking down.
• Use the right channels for reference sources. If a Wikipedia article about your company exists, request corrections on its talk page with independent sources, as its conflict of interest guideline advises.

Aim press work at earning a few strong, independent, accurate stories rather than many thin mentions. The first kind builds the corroborated record that models and reference sources rely on. The second kind, as the Authoritas experiment showed, can leave a name widely published and still absent when people ask about the topic.