PR helps shape what AI understands about a technology category by creating credible third-party information that connects companies with categories, problems, capabilities, use cases and competitors.
A company can say on its website that it belongs to a new category. It can describe the problem it solves, explain why its technology matters and position itself as a leader.
That gives AI systems one source of information: the company itself.
PR expands that information into the public domain. When journalists, reviewers, analysts and industry publications independently discuss the company, they create external evidence that supports, challenges and adds context to its positioning.
At Proper Propaganda, we think of this as publicly corroborated knowledge.
PR helps turn positioning into publicly corroborated knowledge. Publicly corroborated knowledge is easier for AI systems to retrieve, connect, summarize, cite and repeat.
For technology companies, particularly those creating or entering emerging categories, that has become an important function of PR.
Key Takeaways
- PR creates independent evidence about a technology company and its category.
- Earned media can reinforce the relationship between a brand, category, problem, capability and use case.
- Repeated, consistent coverage creates a clearer public record of what a company does and where it belongs.
- PR can contribute to how emerging technology categories are defined and explained.
- Editorial coverage gives AI systems third-party sources they may retrieve and cite when constructing answers.
- PR and GEO increasingly overlap because earned authority can influence AI visibility.
- The goal is accurate, credible and consistent public corroboration of the ideas a company wants to be associated with.
By the numbers
The connection between earned media and AI visibility is becoming much easier to measure.
The following figures come from Muck Rack and Meltwater research into sources cited by major generative AI platforms. The percentages describe their respective datasets and should not be interpreted as universal citation rates across every AI system.
| What the research shows | Why it matters |
|---|---|
| 84% of citations in Muck Rack’s May 2026 study came from earned media | Muck Rack analyzed more than 25 million links from ChatGPT, Claude and Gemini responses across 17 industries. Muck Rack |
| 27% of citations in the same study came from journalism | More than one-quarter of the cited sources were journalistic, showing the role editorial media plays in the AI information ecosystem. Muck Rack |
| 46% of citations for industry-trend queries came from journalism | Muck Rack found journalism was cited at more than twice the rate seen in how-to and comparative queries for this type of question. Muck Rack |
| 37.6% of citation domains in Meltwater’s May dataset were earned/news media | Earned/news remained one of the largest source categories across more than 8 million citations analyzed by Meltwater. Meltwater |
| 0.2% were press releases | In the same Meltwater dataset, press releases accounted for a small fraction of citation domains compared with earned/news media. Meltwater |
| Earned media ranged from 82% to 89% across three Muck Rack studies | The pattern remained relatively consistent across three editions of the research beginning in July 2025. Muck Rack |
The numbers point in the same direction. Independent sources are an important part of the information environment AI systems use to answer questions.
For technology companies, this gives PR a role beyond awareness and reputation. Earned media can create external evidence that helps establish what a company is, which category it belongs to, what problems it solves and where it sits relative to competitors.
That is the basis of what we call publicly corroborated knowledge.
How does PR influence what AI understands about a technology category?
AI systems have a large information ecosystem available to them.
Depending on the system and the query, that can include company websites, media coverage, reviews, forums, academic material, reference sites, industry publications, product pages and many other public sources.
A company’s website is only one part of the picture.
If a technology company describes itself as an AI observability platform, that creates an owned association between the company and AI observability.
Now imagine that technology publications independently describe the company as an AI observability platform, discuss its role in monitoring AI agents, explain how its product traces model behavior and include it in comparisons of AI observability platforms.
A much richer public record now exists:
Company → Category → Problem → Capabilities → Use cases → Competitors
Those relationships are supported by sources beyond the company itself.
This is where PR can affect AI visibility.
What is publicly corroborated knowledge?
Proper Propaganda defines publicly corroborated knowledge as information about a company, product or technology category that is supported by credible independent sources rather than existing solely as a company claim.
Companies control their own positioning. They decide what goes on the homepage, how a product is described and which category name appears in a press release.
The wider market creates another layer of information.
Journalists, reviewers, analysts, customers and other independent sources describe companies from their own perspective.
Consider a few examples:
| Company positioning | Public corroboration |
|---|---|
| “We are an AI infrastructure company.” | Independent technology publications describe the company as an AI infrastructure company. |
| “We created a new technology category.” | Journalists begin using the category terminology and explaining what it means. |
| “Our technology solves X problem.” | Media coverage discusses the product in the context of solving that problem. |
| “Our product is designed for X use case.” | Reviewers test or recommend the product for that use case. |
| “We are a major company in this category.” | Independent comparisons regularly include the company alongside established competitors. |
The right-hand column creates something particularly useful for AI search: external evidence.
How PR shapes AI understanding
The relationship can be summarized fairly simply:
| PR function | What it can establish |
|---|---|
| Category association | What the company is |
| Problem association | What problem it solves |
| Capability association | What the technology does |
| Use-case association | When and why the product is relevant |
| Competitive association | Which companies and products it should be compared with |
| Third-party corroboration | Independent evidence supporting company positioning |
| Category definition | How an emerging technology category is understood |
These associations create the public evidence layer around a company.
There are five areas where PR can have a particularly meaningful influence.
1. PR establishes category associations
One of the first questions an AI system needs to resolve when discussing a company is fairly basic:
What is this company?
For an established brand, the answer may be obvious.
For an emerging technology company, it often is not.
A company might simultaneously be described as an AI company, software platform, developer tool, fintech business, infrastructure provider or agentic commerce company.
The category matters because it affects which questions the company can reasonably appear in.
PR can help establish a more consistent relationship between the brand and the category it wants to occupy.
The important associations usually look something like this:
Brand → Category
Brand → Product
Brand → Problem
Brand → Capability
Brand → Use case
Brand → Competitor
When those relationships appear across credible independent publications, there is more public evidence connecting the company with each concept.
2. Earned media corroborates company positioning
There is no shortage of company claims on the internet.
Every startup can call itself innovative. Every new platform can claim to be transforming its industry. Every founder can create a new category name.
Independent coverage adds evidence and context.
Take a hypothetical payments company.
Its website might say:
“We provide infrastructure for agentic commerce.”
The company has clearly defined its position.
Now imagine that several credible fintech and technology publications independently report that the company provides payments infrastructure that allows AI agents to complete transactions.
The underlying positioning has moved beyond the company’s website.
There is now independent information connecting the company with:
Agentic commerce → AI agents → payments infrastructure → transactions
For GEO, those relationships can be extremely valuable.
Muck Rack’s May 2026 research provides some evidence for why that external layer matters. Across more than 25 million links analyzed, 84% of citations came from earned media and 27% came specifically from journalism. Muck Rack
3. Repeated coverage creates a more consistent public record
Consistency is one of the overlooked benefits of sustained PR.
A single article can introduce an idea. A broader body of coverage can establish the context around it.
Consider an emerging AI observability company.
Over time, independent coverage might establish that:
- the company operates in AI observability;
- AI observability involves monitoring production AI systems;
- the platform traces AI-agent activity;
- evaluation is an important part of the category;
- the technology helps identify failures;
- several specific companies compete in the market.
Those articles do not need to use identical language.
What matters is that the underlying facts and relationships remain consistent.
Eventually the public record becomes much clearer:
AI observability → monitoring + tracing + evaluation → production AI systems → vendors → use cases
There is now a coherent information environment around the category.
This does not mean repetition alone guarantees AI visibility. Relevance, source authority, accessibility and the specific system retrieving the information all matter.
4. PR can help define an emerging technology category
Mature technology categories have decades of public information behind them. There is broad agreement about what a laptop, VPN or CRM platform is.
Emerging categories are messier.
Companies use competing terminology. Definitions overlap. Founders introduce new category names. Analysts use another set of terms. Journalists are trying to explain all of it to readers.
The eventual category definition is shaped by the public conversation that follows.
PR has always played a role in that process.
A good category-building campaign gives journalists enough substance to answer basic questions:
What is this technology?
Why does it exist?
What problem does it solve?
What capabilities define the category?
How is it different from the technology that came before it?
Which companies belong in the category?
Who actually needs it?
As journalists and industry publications begin answering those questions, their explanations become part of the public record.
AI systems now sit downstream from that public conversation.
That gives technology PR another audience.
5. Earned media creates potential source material for AI answers
The relationship between PR and AI visibility becomes clearest when we look at citations.
Generative search systems increasingly provide sources alongside their answers. Those citations provide some visibility into the information being used to support a response.
Journalism is part of that ecosystem.
Muck Rack found that the type of question has a significant effect on what gets cited. In its May 2026 analysis, journalism appeared in 46% of citations for industry-trend queries, more than twice the rate found for how-to and comparative queries. Muck Rack
That is particularly relevant for technology category PR. Questions about emerging technologies, markets and industry changes often sit squarely within this kind of query.
This also makes the quality and relevance of earned coverage more important than raw placement volume.
A passing company mention in an unrelated article may have limited value.
An article that clearly explains what a company does, places it within the correct category, discusses its technology and connects it with a specific customer problem contains far more useful information.
For PR teams, this changes the question from:
How much coverage did we get?
to:
What did the coverage establish about the company?
That is a much better question for both PR and GEO.
A simple example: AI observability
Consider a hypothetical emerging category called AI observability.
Early in the category’s development, the information online might be inconsistent.
One company talks about AI monitoring.
Another talks about LLM tracing.
Another focuses on agent evaluation.
A fourth uses the term AI reliability.
Someone asking an AI system “What is AI observability?” may receive a broad or inconsistent answer because the public information around the category itself is still developing.
Now imagine a year of meaningful industry coverage.
Technology publications begin explaining AI observability around several recurring ideas:
- monitoring AI applications;
- tracing AI-agent actions;
- evaluating model performance;
- identifying failures in production;
- understanding how AI systems reach outcomes.
Those publications also begin consistently identifying several companies as AI observability vendors.
The public understanding of the category becomes more structured.
| Element | Public understanding |
|---|---|
| Category | AI observability |
| Core function | Monitoring production AI systems |
| Capabilities | Monitoring, tracing and evaluation |
| Problem | Understanding and diagnosing AI behavior |
| Environment | Production AI applications and agents |
| Vendors | Companies consistently associated with the category |
| Use cases | Reliability, debugging, evaluation and performance monitoring |
No single PR placement created that understanding.
The accumulated public record did.
PR can play a significant role in creating that record.
Does PR literally teach AI models about a company?
Publishing a media story does not mean ChatGPT immediately “learns” the information in the article.
AI products use different models, indexes, retrieval systems and data sources. Those systems also change.
For PR purposes, the more useful way to think about the relationship is through the information environment.
PR creates credible public information.
Search engines can index it. Web-connected AI systems can retrieve it. Generative search systems can use web sources to support or contextualize answers.
Google, for example, says its generative AI Search features are rooted in its core Search ranking and quality systems and use techniques including retrieval-augmented generation to retrieve relevant, current pages from its Search index.
For communications teams, the practical implication is straightforward:
The clearer and more credible the public information surrounding a company and its category, the better the source environment available to systems trying to understand it.
Is earned media more valuable than a press release for AI visibility?
Press releases still have a job.
They establish dates, announcements, product specifications, executive comments, company descriptions and other official facts.
Earned media performs a different function.
A journalist can explain why an announcement matters, place the technology within a market, compare it with alternatives, question the company’s claims and describe the product in independent language.
That makes editorial coverage particularly useful for public corroboration.
Current citation data shows a substantial difference between the two as direct AI sources.
Meltwater’s June 2026 AI Search Visibility Report analyzed more than 8 million citations across eight major LLMs. Earned/news media represented 37.6% of citation domains in May 2026, while press releases represented 0.2%. Meltwater
| Source | Role in the information environment |
|---|---|
| Company website | Establishes owned facts, products and positioning |
| Press release | Provides official announcements and source information |
| Earned media | Adds independent corroboration and context |
| Product review | Connects products with performance, features and use cases |
| Comparison article | Establishes category and competitive relationships |
| Executive interview | Connects people and companies with expertise |
| Industry analysis | Helps explain categories, markets and trends |
A strong GEO information environment will usually contain several of these.
The data does not make press releases irrelevant. Their value may also be indirect. They provide structured source information that can lead to reporting and other third-party coverage.
The important distinction is between publishing a company claim and earning independent authority around it.
PR’s role in technology category creation
Category creation has always depended heavily on language.
Someone names the category.
Companies begin using the term.
Journalists decide whether it is useful.
Analysts refine the definition.
Competitors adopt it, modify it or fight it.
Customers eventually start using it themselves.
Generative AI has added another layer because people now use AI systems to research products, companies, markets and unfamiliar technologies.
Questions might include:
- What is agentic commerce?
- What is AI observability?
- What is embodied AI?
- What is an AI PC?
- What are smart rings?
- Which companies make AI-powered wearable devices?
- Who are the leading companies in this category?
Technology companies are competing for inclusion in those answers.
They are also competing over the definition of the category itself.
For companies trying to establish a new market, that makes category PR and GEO increasingly difficult to separate.
What should technology PR teams do differently for AI search?
The starting point remains the same: create stories people actually want to read.
Journalists do not care about your GEO strategy, nor should they.
The change happens behind the scenes.
Communications teams need to become more deliberate about the facts, terminology and relationships they are trying to establish in the public record.
At Proper Propaganda, we think that starts with eight questions:
- What category do we want the company associated with?
- How should that category be defined?
- Which customer problems should the company be connected to?
- Which capabilities require independent corroboration?
- Which use cases have the greatest commercial importance?
- Which competitors does AI currently associate with the category?
- Which publications and sources are being cited in relevant AI answers?
- Where does the company’s positioning differ from how the wider web describes it?
Answer those questions and PR becomes much more precise.
You know which ideas need public support, which narratives are poorly established and which sources matter to the conversations you want to enter.
This is also why relevance matters alongside raw media reach. Muck Rack’s technology-specific analysis found AI systems collectively citing journalism from more than 20,000 different publications, spanning trade, local and national media. No single publication dominated the citation landscape. Muck Rack
Where PR fits into the Proper Propaganda GEO Framework
At Proper Propaganda, we don’t treat PR as a separate GEO tactic.
It sits inside our broader Generative Engine Optimization framework, which has six phases:
Audit → Strategy → Foundation → Content → Authority → Measurement
The framework is designed to improve how a brand is understood, cited, represented and recommended by AI systems.
PR plays its biggest role in Authority, but its impact starts much earlier.
1. Audit: Understand how AI currently sees the category
Before trying to change the public narrative, you need to know what already exists.
For a technology company, that means looking at:
- how AI systems define the category;
- which companies they associate with it;
- which competitors appear most often;
- which capabilities and use cases are commonly mentioned;
- which publications and websites are being cited;
- where the company’s own positioning differs from the wider public record.
This establishes the existing information environment.
If a company wants to become associated with “agentic commerce,” for example, the first question is what AI currently understands agentic commerce to mean, which companies it associates with the term and which sources are shaping those answers.
2. Strategy: Decide what you want to be known for
The Strategy phase determines which categories, topics, prompts, audiences and business outcomes matter.
This is where positioning becomes important.
A technology company needs to decide which associations it wants to establish:
Brand → Category → Problem → Capability → Use case → Expertise → Competitors
Those relationships give the PR program something concrete to build around.
For category-building companies, Strategy also means deciding how the category itself should be explained.
What does the term mean?
What problem created the need for it?
Which capabilities define it?
What technologies sit outside the category?
Which companies belong in it?
Those are PR questions, but increasingly they are GEO questions too.
3. Foundation: Make the owned information clear
The website needs to give search and AI systems a clear explanation of the company before PR starts creating external evidence around it.
Our GEO framework calls this the Foundation.
That includes site architecture, internal linking, structured data, entity optimization, technical SEO and crawlability.
It also includes something much more basic: consistency.
The company’s website should clearly explain:
- what the company is;
- which category it belongs to;
- what its products do;
- which problems they solve;
- who they are designed for;
- what makes them different.
The language used in PR should connect back to that foundation.
4. Content: Create information worth retrieving
The Content phase creates owned material that explains the category and gives search and AI systems useful information to retrieve.
Educational articles can define terminology.
Original research can establish facts and statistics.
Comparison pages can explain how technologies differ.
FAQs can answer specific questions.
Product documentation can clarify capabilities and use cases.
This content also gives PR something substantive to work with.
A strong PR program should have useful information, research, expertise and evidence to bring into the wider conversation.
5. Authority: Turn positioning into publicly corroborated knowledge
This is where PR has its most direct impact.
Our GEO framework defines Authority as the process of building external signals that increase trust in the brand.
That includes:
- earned media;
- digital PR;
- expert commentary;
- thought leadership;
- third-party citations;
- reviews;
- relevant industry conversations.
This is where positioning can become publicly corroborated knowledge.
The company has established what it wants to be known for. Its website and content explain those ideas clearly. PR creates opportunities for credible independent sources to examine, describe and contextualize them.
Over time, those external sources can establish a broader public record:
Company → Category → Problem → Capabilities → Use cases → Expertise → Competitors
The company no longer stands alone in making those associations.
Other sources support them.
6. Measurement: Find out whether AI understanding actually changed
Coverage itself does not tell you whether the work succeeded.
Measurement tells us whether Foundation, Content and Authority are actually changing how AI systems represent the brand.
| Metric | What it tells us |
|---|---|
| Share of Answer | How often the brand appears across important AI prompts relative to competitors |
| Brand Citation Rate | How often the brand or its owned properties are cited |
| Prompt Coverage | How much of the commercially important prompt universe includes the brand |
| Product Ranking | Where products appear in AI-generated recommendations |
| Share of Shelf | How visible the brand’s products are within AI-generated recommendations |
| Citation Sources | Which publications and websites are influencing relevant answers |
| Narrative Accuracy | Whether AI systems describe the company, products and capabilities correctly |
| Category Association | Whether AI systems consistently place the company in the desired category |
| Recommendation Strength | Whether the brand is simply mentioned or actively recommended |
| Sentiment | How positively or negatively the brand is represented |
Measurement then feeds back into the rest of the program.
If AI systems associate the company with the wrong category, revisit Strategy, Foundation and Authority.
If competitors dominate because AI repeatedly cites particular publications, those publications become Authority opportunities.
If the brand appears frequently but its differentiation is poorly understood, the problem may sit in Content, PR messaging or both.
This is why our GEO framework is a loop rather than a six-step campaign.
How PR moves through the GEO framework
| GEO phase | Role of PR |
|---|---|
| Audit | Identify the existing category narrative, competitors and sources influencing AI answers |
| Strategy | Define the category, problems, capabilities and associations the brand wants to own |
| Foundation | Make sure owned positioning and brand information are clear and consistent |
| Content | Develop useful information, expertise and evidence worth citing |
| Authority | Build independent third-party corroboration through earned media and external sources |
| Measurement | Determine whether those signals are changing AI visibility, citations and understanding |
The process creates a logical progression:
Understand the existing information environment → Decide what you want to be known for → Make that information clear on owned properties → Create useful evidence and content → Build independent authority around it → Measure how AI responds
PR is especially important in the transition from what a company says about itself to what the wider information environment says about the company.
That is the public evidence layer.
How we developed this analysis
This analysis combines Proper Propaganda’s work measuring brand visibility across generative AI platforms with published research into the sources AI systems retrieve and cite.
We look at AI visibility through our six-stage GEO framework:
Audit → Strategy → Foundation → Content → Authority → Measurement
For PR specifically, we examine how owned positioning is reflected across independent media sources and whether those associations subsequently appear in AI-generated answers.
We look at:
- Share of Answer;
- Brand Citation Rate;
- Prompt Coverage;
- Citation Sources;
- Narrative Accuracy;
- Category Association;
- Product Ranking;
- Share of Shelf;
- competitive positioning.
External evidence cited in this article includes Muck Rack’s research into millions of AI citations, Meltwater’s analysis of citation sources across major LLMs and first-party documentation from major search platforms where applicable.
The purpose is not to suggest that PR coverage directly changes an AI model. It is to understand how PR changes the public information environment available to systems retrieving information about a company or category.
In one sentence
PR influences AI visibility by creating independent public evidence that connects a company with the categories, problems, capabilities and use cases it wants to be known for.
The Proper Propaganda view
PR has always influenced how markets understand companies.
Now some of that understanding is being mediated by machines.
A prospective customer might ask an AI system:
What is this technology?
Which companies make it?
What problem does it solve?
Who are the leaders?
How does Company A compare with Company B?
Which product should I buy?
The quality of those answers depends, in part, on the information available to the system.
Companies control some of that information.
The wider market creates the rest.
That is why earned media matters.
PR helps turn positioning into publicly corroborated knowledge. Publicly corroborated knowledge is easier for AI systems to retrieve, connect, summarize, cite and repeat.
For technology companies, particularly those trying to establish a new category, that makes PR part of the infrastructure of AI visibility.
Everything you need to know about PR and AI category understanding
| Question | Answer |
|---|---|
| How does PR shape what AI understands about a technology category? | PR creates credible third-party information connecting companies with categories, problems, capabilities, competitors and use cases. |
| Does PR directly teach AI models? | Publishing coverage does not immediately update an AI model. PR changes the public information environment that web-connected search and AI systems can retrieve from and use as evidence. |
| What is publicly corroborated knowledge? | Proper Propaganda defines it as information about a company, product or category that is supported by credible independent sources rather than existing solely as a company claim. |
| Why does earned media matter for GEO? | Earned media provides independent evidence and context that can become part of the source landscape available to AI systems. |
| Can PR help create a technology category? | Yes. Media coverage can help establish the terminology, definitions, problems, capabilities, examples and companies associated with an emerging category. |
| Does more coverage automatically improve AI visibility? | No. Relevance, authority, context and consistency matter. |
| Are press releases useful for GEO? | Yes, particularly for official facts and announcements. Earned editorial coverage adds independent context and corroboration. |
| How do PR and GEO work together? | PR creates external evidence and authority. GEO measures and improves how the brand is surfaced, described, cited, compared and recommended in AI answers. |
| What should technology companies measure? | Share of Answer, Brand Citation Rate, Prompt Coverage, Product Ranking, Share of Shelf, Citation Sources, Narrative Accuracy and Category Association. |
FAQs about PR, GEO and AI search
How does PR help AI understand a technology company?
PR creates independent information about what a company does, which category it belongs to, the problems it solves and the capabilities associated with its products. Consistent coverage across credible sources creates a stronger public evidence base around the company.
Why is earned media important for AI search?
Earned media provides independent information about brands and categories that can become part of the source landscape available to search and AI systems. Muck Rack’s May 2026 analysis found earned media represented 84% of citations within its dataset of more than 25 million links from ChatGPT, Claude and Gemini responses. Muck Rack
Can PR influence what ChatGPT says about a company?
PR can influence the public information environment available to web-connected AI systems. It cannot guarantee a particular ChatGPT answer. Clear owned information, relevant earned media and consistent external corroboration can strengthen the evidence available about a company.
Can PR help establish a new technology category?
Yes. PR can introduce category terminology, definitions, problems, capabilities and use cases into wider public discussion. When independent journalists and industry sources begin using those concepts, they become part of the public record around the category.
How much media coverage is needed to influence AI visibility?
There is no universal number. Relevance, authority and context matter more than hitting an arbitrary placement target. The objective is to build a credible and consistent body of independent information around the company and the subjects it wants to own.
Are press releases as useful as earned media for AI visibility?
They serve different purposes. Press releases establish official facts and announcements. Earned media provides independent context and corroboration. Meltwater’s May 2026 dataset found earned/news media represented 37.6% of citation domains, compared with 0.2% for press releases. Meltwater
What is the difference between PR and GEO?
PR builds public visibility, authority and third-party corroboration. GEO focuses on how that information translates into visibility within generative AI systems, including whether a company is surfaced, cited, accurately described, compared and recommended.
What is the most important PR principle for AI visibility?
Build accurate, consistent and independently corroborated associations between the company and the categories, problems, capabilities and use cases it wants to own.