PR ON THE GO • June 15, 2025

Why AI-Generated Content Won't Get You Far in the Press

We asked our PR and growth experts to advise our startups and creative entrepreneurs on using AI in media relations. Read on to find out why AI-generated press releases, expert submissions, and interview responses are usually not published when pitching for press coverage.

Our experts explained how AI content detectors, such as GPTZero, prevent AI-generated pitches from being considered.

Combating Disinformation in the Age of Artificial Intelligence: The experts provide examples of AI hallucinations in media pitches and explain how AI fabricates current context and experiences. They also offer advice on best practices, such as using human-written content with AI-assisted grammar corrections and leveraging an entrepreneur's authentic voice.

• AI generates lowest-common-denominator writing
• Authentic first-hand experience is key
• Editors are looking for a story that feels lived in
• In marketing, trust is all you have
• Editors can sense when a submission feels soulless or rehearsed
• AI lacks coherence
• AI creates “faceless” content
• Leverage AI to analyze data, but don't use it to write your pitches.
• Reporters need knowledge that you have gained through your own experience
• Most AI-generated press releases read like a LinkedIn post on life support.
• Google's algorithm is now discounting "predictable language"
• Hallucinations erode credibility immediately
• If you let AI talk for you, you sound like everyone else.
• Generic phrasing, vague claims, or a lack of specific lived experience
• Speed matters, but believability matters more
• AI generates past collaborations that simply never happened
• If the heart of your story isn’t yours, it’ll show.
• AI adds events that never happened
• AI completely fabricates context or experiences
• AI content lacks the true human experience and voice that media seeks
• Journalists notice the repetitive patterns of AI
• AI misrepresentation will destroy the valuable trust with the media
• Journalists want real-world knowledge than AI can't supply
• Journalists receive AI-generated press releases full of hallucinated false facts
• AI can’t genuinely respond to the news cycle
• AI-generated press content isn't trusted
• Media pitches require human involvement
• AI misses your lived experience
• Don't get flagged as synthetic content
• AI tends to produce overly tidy but generic fluff
• The core story must come from the founder, not AI

AI generates lowest-common-denominator writing

Emily Reynolds-Bergh, Owner at R Public Relations

"Even before AI, it was difficult to stand out in a sea of email pitches. Now, it’s more important than ever to write unique pitches. AI tends to generate lowest-common-denominator writing, meaning it uses a lot of platitudes and vague language to fill gaps. As a writer, PR expert, or entrepreneur it’s your job to find a concise, catchy way to share precise messaging. Your pitch doesn’t need to be long or flashy; it just needs to feel real and authentic. However, I am a big fan of AI grammar tools like Grammarly, which help refine human-written content."

Authentic first-hand experience is key

Paige Arnof-Fenn, Owner at Mavens & Moguls

"With all the AI tools generating robotic messages that sound generic, your brand can stand out and break through the sea of sameness with personalized pitches, thoughtful communication serving your audience's specific needs. Perplexity, Claude, ChatGPT are great tools that can help you generate ideas and brainstorm but they cannot compete with or replace your humor, experience or war stories.

Authentic first-hand experience is a key trend in SEO today, Google has said for years that experience, expertise, authority, and trust matter to qualify but the prevalence of AI-generated content and users’ search for authenticity have made those qualities critical for anyone pitching and trying to rank content now.

When you share what you know -- your passion, your war stories, the good, bad and ugly -- the stories will be interesting and the lessons will be real, people will remember you and come back for more. Tell your origin story, share mistakes/failures, be vulnerable if you want to drive engagement, build media connections and relationships with your audience and show your humanity which is more important for now than ever before! To build trust you have to connect on a personal level."

Editors are looking for a story that feels lived in

Pilar Lewis, Public Relations Associate at Marketri

"Journalists can tell when something’s been written by a machine. There’s a generic sound to it, often repeating the same verbiage. It like it’s trying to say the most professional, though out phrases without actually holding meaningful substance. I’ve seen pitches that sound impressive on the surface, but that is because they’ve been perfectly crafted by a LLM rather than holding human context.

Editors aren’t just looking for clean copy; they’re looking for a story that feels lived in. And when something reads like it could’ve been written about any company, it’s a missed opportunity to connect.

I've seen AI-generated content include details that simply aren't grounded in reality and that’s a quick way to lose credibility with a journalist. They don’t want to read the same generic quote. They want transparency. They want to understand who you are, what you’re building, and why it matters right now.

AI can absolutely help with polishing your message, but it can’t replace your voice. The best press pitches come from a place of real experience: what you’ve struggled through, what you’ve learned, what you’re excited about. That’s what gets people to lean in."

In marketing, trust is all you have

Linda Yang, Co-Founder & Digital Marketing Strategist at TotePrint.com

"One aspect of my job is assisting brands in telling real, impactful stories, particularly when they are positioning themselves for media coverage. Let’s discuss press releases and interviews, then.

Why AI-Written Press Pitches Usually Don’t Work
You may be thinking, “Aha, I’ll just have AI generate my press release.” Well, here’s a secret: reporters know if it’s written by a robot. Sounds too formal. Too generic. Lacks your actual voice. That’s trouble, since reporters want to hear from you, you, actual person behind the company.

There are also some specific tools (referred to as AI content detectors) that check what you send. If your message appears as if it has originated from AI, it could be deleted before anybody even reads it. Yes, that quickly.

AI Sometimes Makes Stuff Up (Seriously)
AI does come in handy, but it does fabricate, as well. That is an “AI hallucination.” I’ve had it tell me a company won an award they weren’t even nominated for, or they were included in a magazine they weren’t. That damages your credibility with the press. And in marketing, trust is all you have.

What Actually Works
This is what I advise my customers, and what I practice myself:

• Write it yourself using my actual story and words.
• Use AI to fix grammar or spelling, not to write the whole thing
• Be truthful regarding what you've accomplished and what you've learned.
• Sound like a person, not a machine

When I pitch our brand narrative on TotePrint.com, I always stay honest. We share how we began, what issues we address, and how our merch is different. People respond to that. Editors respond as well.

Final Tip: You Are the Best Part of Your Brand.
You don’t have to use big words in order to be smart. You just need your words to be real. That’s what differentiates your brand. Your story is strong, just ensure it’s yours. You've got it!"

Editors can sense when a submission feels soulless or rehearsed

Angel Sanchez, Owner & Photographer at Wanderlust Portraits

"The audit process for AI-generated results starts with checking what LLMs are surfacing about your brand. That means searching your company name, products, leadership, and key phrases in AI tools and reviewing the summaries they produce. Cross-check those results with earned media, owned content, and third-party sources. You’re not just looking for misinformation, you’re looking for gaps. If your best product story is missing, or an old price point is still showing up, that tells you where the work needs to start.

The future of brand presence in AI-generated search will depend on clear, consistent earned coverage. LLMs weigh reputable sources like news articles and government sites more heavily than your blog. That means PR teams are not just promoting, they’re anchoring your truth. If a new product gets three solid pieces in respected publications, you increase the chance that AI will repeat it accurately. The same applies to executive thought leadership, product recalls, and pricing changes.

At EcoATM, we focus our PR on stories that drive user confidence and product clarity. When a local outlet covers how our kiosks help reduce e-waste, that’s not just a win for brand visibility, it’s an input that AI can repeat to users asking where to recycle phones."

AI lacks coherence

Jeff Romero, Founder at Octiv Digital

"I have noticed that there is a lot of poorly written content on the internet. AI-powered systems curate information from existing web sources. This raises the likelihood that it may produce mediocre content. Researchers see that GPT-3 proceeds in a linear manner. It does not rewrite work in the same manner that a human does. The completed product may contain gibberish or lack coherence. Worse, current research indicates the longer the content, the more errors it produces. These platforms cannot successfully fact-check work since it frequently requires specialized knowledge.

Finally, professional writers improve their skills with time. With experience, they may create a unified, fluid piece of material for a specific audience in the appropriate voice and tone."

Leverage AI to analyze data, but don't use it to write your pitches.

Jon Kelly, SEO Specialist & Owner at LinkBuilding.co

"I ask clients for feedback and get them to fill a pretty in depth questionnaire to get a full profile on the client so that when relevant queries come up, their true expertise and experience can be reflected in the pitch.

Regarding AI, the detectors are getting more efficient by the week - from analyzing sentence structure to the frequency and probability of words used, depending on AI for your pitches is not a good use of time. Instead, leverage AI to analyze and summarize data, questions, etc. and then use that data in your pitches. Being your true human self will be obvious more often than not in pitches."

Reporters need knowledge that you have gained through your own experience

Marcus Denning, Senior Lawyer at MK Law

"With what I see, AI technology fails to provide content that reporters look for. I have experience dealing with various property disputes, immediate family law cases, and business refinements, and details are crucial in each of them. When giving a comment, I rely on my training, and journalists can tell it is different from someone’s view from afar.

I feel like the way people interact with each other counts more than their words. Reporters believe in you because you talk honestly and only share knowledge that you have gained through your own experience. Without trust, AI will not cut it in the world of media, so your efforts will be limited.

In addition, reporters are interested in getting beyond the basic description of the news. When there is no personal connection to this kind of work, it usually goes unnoticed and is easily ignored.

For me, you earn trust and credibility with people, not by just copying answers under a computer’s guidance. If you wish to gain valuable insights, it must be shared by someone who has got hands-on experience. Here, nothing can be achieved without hard work."

Most AI-generated press releases read like a LinkedIn post on life support.

Paul Allen, Growth Strategist at DublinRush.com

"Most AI-generated press releases read like a LinkedIn post on life support. They’re technically sound, sure—but dead on arrival.

No pulse. No perspective. No person.

Editors don’t reject AI content because they’re scared of the robots.

They reject it because it’s boring.

Here’s what actually happens:

Founders delegate the press release to ChatGPT.

It strings together generic phrases like “innovative platform revolutionizing XYZ,” hallucinating fake milestones and fabricating context that sounds real enough to pass a skim—but collapses under scrutiny.

One pitch I saw referenced a Forbes interview that never happened.

That’s not just a mistake—that’s a trust nuke.

AI detectors are the least of your worries.

The human detector is the one that kills your story.

Editors and journalists can feel it when a pitch wasn’t lived—it was just typed.

Instead, use AI like a co-pilot.

Let it fix your grammar. Let it reword clunky paragraphs.

But the voice? The scars? The “I failed twice before I figured this out” story? That’s all you.

In 2025, the new luxury isn’t automation—it’s authenticity.

Pitch like a person, not a prompt. That’s how you earn ink."

Google's algorithm is now discounting "predictable language"

Devajit Mutsuddy, PR Specialist at Pearl Lemon PR

"Whatever strengths AI has will fade quicker than a toddler spots an iPad. And the biggest reason is that AI simply lacks the lived nuance and contradictions at the heart of human stories. From tone-deaf intros and mind-boggling hallucinations (we once received a pitch from a start-up founder where they claimed scale Mt. Everest and start a unicorn in the same year (they did neither)), AI convincingly fabricates context, and that's deadly in PR.

AI detectors are becoming better - and more nuanced - so Google's algorithm is now discounting "predictable language," not to mention synthetic tonality. Editors want uninhibited brilliance, not rigidly structured nonsense. At Pearl Lemon PR, we've had instances of clients getting ghosted by the media only until we released pitches telling stories of their grit - one founder even launched a company while couch surfing in Brixton.

The best? We use AI. Not to write the story, but to clean up grammar, ensure facts have been checked, and create cohesion of voice. It’s like using spellcheck on the computer, not Shakespeare. The soul still needs to come from the founder."

Hallucinations erode credibility immediately

Eugene Leow, Director at MarketingAgency.sg

"Journalists can smell AI-generated pitches a mile away. They’re vague, overly polished, and full of generic claims. We once tested an AI-only draft for a pitch and got zero responses. Why? It lacked specifics, cited non-existent data, and made up a “recent product launch” that didn’t happen. These hallucinations are dangerous; they erode credibility immediately. Plus, many outlets now run content through detectors, if your quote flags as synthetic, it often doesn’t make it past the inbox.

Start with your own story. We help founders draft with their voice then use AI for minor grammar polish, not substance. A good pitch doesn’t sound perfect. It sounds real. Editors want quotes they can believe, not blurbs that sound like marketing copy dressed up by a machine."

If you let AI talk for you, you sound like everyone else.

Thomas Franklin, Founder & CEO at Swapped.com

"I know how press teams respond to noise. And honestly, AI-generated PR pitches? They are noise!

AI press releases get tossed fast because they sound like Wikipedia with an ego problem.

Every pitch starts sounding like “XYZ disrupts the market with cutting-edge innovation.” No reporter on earth is excited to read that. The writing rings hollow, the quotes are lifeless, and worse, half the time the tool hallucinates fake partnerships or made-up growth stats. I once read a pitch that said a company "partnered with NASA", total fiction, just inserted by the model. If that pitch made it through, it could tank credibility for everyone involved.

AI content detectors are not just grammar bots anymore.

They flag structure, rhythm, and lack of emotional variation. I mean, if your quote sounds like it came from a LinkedIn template, the detector scores it low and filters it out. Most outlets now use these tools in their inboxes. That means if you write with generic phrasing, you are toast before a human even reads it. Real people speak in uneven bursts, not perfect cadence. The press wants quotes, not polished blurbs. Big difference.

So yeah, write your pitch yourself!

Use AI to smooth it, not shape it. Get someone to run grammar over it, but let the phrasing, tone, and story come from your own voice. Say something risky. Mention numbers, mess-ups, pivots. The human stuff is what gets printed. AI does not know your path, your market, or your mindset. If you let it talk for you, you sound like everyone else. And if you sound like everyone else, nobody cares.

Real stories cut through. AI writes summaries. Reporters do not publish summaries, they publish stories."

Generic phrasing, vague claims, or a lack of specific lived experience

Kaz Marzo, Operations Manager at Image Acquire

"I’ve worked on both sides of the pitch—crafting outreach and receiving it—and I can tell you from experience: journalists are not just looking for information, they’re looking for human connection. That’s why AI-generated press releases or interview responses almost never get picked up. They often miss the mark emotionally, contextually, and factually.

Most newsrooms now run submissions through AI-content detectors, and anything flagged as synthetic is either deprioritized or outright rejected. These tools are surprisingly accurate at identifying generic phrasing, vague claims, or a lack of specific lived experience. For example, I’ve seen AI pitches that claim a founder was “named to Forbes 30 Under 30” when they weren’t, or that reference speaking at events that never happened—classic AI hallucinations that undermine credibility instantly.

The best practice? Start with your real voice. Use AI to tighten grammar or structure, but the content itself must reflect personal insight, recent activity, and real wins. At Image Acquire, we use AI to polish, never to fabricate. Authenticity builds trust. Journalists can tell when your pitch isn’t coming from lived experience—and so can their readers."

Speed matters, but believability matters more

Mark Friend, Company Director at Classroom365

"In growth and PR, speed matters, but believability matters more. AI-generated press materials often flunk that second test. Editors and journalists are now filtering out pitches that carry the telltale traits of synthetic copy: generic phrasing, vague timelines, and fictitious accomplishments. Some even run content through detectors that flag AI-generated structure, making it easy to bin cold emails before a human sees them.

The challenge goes deeper than looks. AI tends to hallucinate filling in blanks with fake partnerships, made-up funding rounds, or product capabilities that don’t exist. I’ve seen founders submit pitch decks where the AI decided they’d already raised a Series A when they hadn’t even opened a SAFE.

On top of that, here’s what works better like have a human write the core message, something grounded, concise, and specific to the moment. You can run it through a language tool afterward for tone or grammar, but the structure and story should be yours. When I advise startup founders on technical messaging, I remind them that PR isn’t about shouting the loudest. It’s about saying one thing that only you could say. And saying it clean."

AI generates past collaborations that simply never happened

Caspar Matthews, Director at ElectComm Group Electrical and Data

"When it comes to getting your startup or creative venture noticed in the press, you absolutely cannot rely on AI to write your press releases or interview responses. I've been in the electrical field for nearly twenty years, and I can tell you firsthand that real connections and genuine communication are what open doors, not some algorithm.

Here’s why those AI-generated pitches usually hit the digital bin without a second glance. The folks in media have tools, these "AI content detectors," that are pretty good at sniffing out anything that doesn't sound human. Think about it. These detectors are built to flag patterns that aren't natural, those repetitive phrases, the lack of real emotion, or the way an AI sometimes strings words together without genuine understanding. It's like trying to pass off a mass-produced, flimsy light fitting as a custom-built, high-quality piece of electrical work. You just know it’s not the real deal when you see it.

I've even seen some pretty wild "hallucinations" come out of AI-generated content. A while back, I saw a pitch from a startup that had clearly used AI. It mentioned a breakthrough in "quantum flux capacitors" for home energy, and then went on to describe a past collaboration with Electcomm Group on a massive smart home project that simply never happened. The AI had just pulled together a bunch of impressive-sounding words and fabricated a context and an experience that had no basis in reality. It was a complete fiction, and it immediately made the whole pitch seem unreliable. That kind of fabrication, where AI just invents experiences or current contexts, completely undermines your credibility. How can a journalist trust you if your own story isn't true?

What you need to do is write your own content. Pour your heart and soul into explaining what you do, why it matters, and what makes your business tick. You can use AI for grammar checks, sure. Think of it like using a spirit level. It helps you get things straight, but it doesn't do the actual building. The real value is in your own authentic voice. That genuine passion, that personal touch, that's what resonates with people. It’s what makes a journalist want to hear more, because they know they’re getting information straight from the source, from someone who truly understands their own business. It's the difference between a perfectly installed, compliant electrical system and something slapped together that might cause problems down the line. Trust me, the human touch always wins."

If the heart of your story isn’t yours, it’ll show.

Matthew Tran, Engineer & Founder at Birchbury

"Trying to get a story in the news means more than just delivering a formal press release. Journalists search for genuine stories that appeal to and interest their readers. I discovered this when Birchbury was launched. The first press release I made was reliable but did not reflect what we were about. When we decided to share the real story of reaching $1 million in sales in one year by tuning in to customers’ needs and improving our product, news outlets noticed us. They want to understand your reasons for making the decisions you do, and not only the facts.

I’ve realized that AI-generated content can miss that personal touch. You can use it to clean up grammar or tweak the structure, but if the heart of your story isn’t yours, it’ll show. I remember reading a pitch once where it sounded like they were claiming to “revolutionize” their industry, but it didn’t make sense to me. There were no real stories or data behind it. When you’re pitching, make sure the emotion, challenges, and real moments are in there. That’s what makes your story stand out to a reporter."

AI adds events that never happened

Moti Gamburd, CEO at Raya's Paradise

"We’ve earned coverage through real stories grounded in our team’s daily experience, not automated claims. That difference shows up clearly to editors and reporters.

AI-generated press materials often fail because they sound polished but empty. We once tested an AI draft for a caregiver spotlight pitch. It added an event that never happened and quoted a resident who didn’t exist. Those hallucinations happen when AI tries to fill in gaps with assumed details. Even if the grammar is clean, the facts break trust.

We’ve had better success using real stories and refining them with AI for tone and clarity, not for substance. Editors respond to authenticity, a founder’s actual voice, team anecdotes, or a clear outcome from a program. Use AI like spellcheck, not like a ghostwriter. Truth and detail are what make a pitch land."

AI completely fabricates context or experiences

Tracie Crites, Chief Marketing Officer at HEAVY Equipment Appraisal

"When it comes to why AI-generated press releases and interview responses are usually not published, it boils down to two main things: authenticity and nuance. Journalists and editors are looking for a unique story, a real human voice, and genuine insights. At the present stage of AI's capabilities, it will not be able to do this. Press releases and interview answers are often required to reflect a level of excitement, urgency, or comprehend the pain points of an industry; AI-generated content typically represents a bland summary of facts and doesn't have that emotional, impactful storytelling. AI can simulate human language, but it does not truly understand the context, the emotional appeal, or a compelling perspective that tells a story.

One big reason these pitches aren't considered is that AI content detectors are getting pretty sophisticated. These tools measure a range of different characteristics with writing style, sentence structure, predictability of words, and "burstiness," which refer to variation in sentence complexity and length. AI-generated content typically displays less "perplexity" because generative writing is more tied to common language patterns and more predictable, which makes the content feel more uniform. Human writing is more likely to have unexpected choices of words, and unexpected patterns in sentence length, so it has a varied rhythm. Journalists are likely using these detectors (or they might just have a very developed "gut feeling" for AI-generated text), so they will quickly realize they have received a pitch or other communication that lacks that human touch.

I've seen examples of AI hallucinations in media pitches where the AI completely fabricates context or experiences, and it's a huge problem. It's possible to receive a pitch that claims, "Your startup recently made a partnership with Acme Corp to reinvent the widget industry," when Acme Corp does not exist or you've never spoken with them. I have seen cases where an AI system created a quote from a founder about a "groundbreaking study" that never happened; or suggested non-existent awards. AI models can generate false ideas that sound plausible, as they will only predict the next most likely word from their training information, not because they fact-check or verify. This totally incomprehensible aspect makes it impossible for a journalist to trust the pitch at all.

My advice for good practices is to absolutely prioritize human-written content that truly captures the entrepreneur's authentic voice. Use AI tools for things like grammar correction, proofreading, or perhaps brainstorming broad ideas, but the core narrative, the unique selling proposition, and the personal anecdotes should always come from a human. Your authentic voice is what will differentiate you; it’s your personality, your passion, and your unique perspective that no AI can replicate. Let your genuine enthusiasm for your business shine through. That's what will resonate with journalists and help your story stand out in a sea of generic pitches."

AI content lacks the true human experience and voice that media seeks

Mariana Delgado, Marketing Director at DesignRush

"AI-generated press releases and interview responses are often unreleased because they lack the nuance and authenticity that journalists like to see. AI content lacks the true human experience and voice that media outlets often prioritize. Even though AI content is great for efficiency, it often loses emotional depth, misses real-world context, and can come across as too generic and robotic. This focus on the voice of the pitch is what ultimately makes a pitch resonate with a reporter."

AI content detectors are getting better at recognizing patterns that seem to be associated with AI-generated content, and non-indicative inconsistency. Specifically, these detectors identify distinguishing factors such as clichéd phrasing, repetitive sentence structure and overly formal language common AI text characteristics; recognizing that all can be a warning sign of non-authenticity or unreliability. The danger of being flagged often causes an AI-generating pitch to be seen as non-authentic before it even makes it to the reviewer.

AI hallucinations can happen in media pitches when the tool generates information that is factually inaccurate or include outright fabrications. For example, AI may generate a pitch that states a start-up has been "featured" in a publication that it hasn’t, or the AI could generate statistics or industry insights that sound possible but are actually made up. Often AI has difficulty incorporating context from real world experiences, resulting in pitches that get the dates wrong, or the location wrong, or even the company's history wrong. Journalists recognize these inaccuracies quickly, and they undermine the credibility of the pitch.

For best practices, entrepreneurs should use AI to enhance, not replace, authentic content. Entrepreneurs focus on writing in their own voice, and that voice resonates with their audience. Then the entrepreneur can use AI just for grammar and spelling corrections or to optimize for understanding, but the essence of the message must come from the real leader of the business. Using an entrepreneur's authentic voice in a media pitch gives the pitch a more compelling, personal, and trustworthy narrative something that AI cannot duplicate yet."

Journalists notice the repetitive patterns of AI

Kristiyan Yankov, Co-founder & Growth Marketer at Above Apex

"Anyone who has ever worked on content creation can notice AI-written content from a mile away. To put it simply – it looks like another page from Wikipedia, with some out-of-place clichés here and there to imitate real human communication. After a while, you start to notice the patterns, which are really repetitive.

What drives a lot of the press coverage and PR is providing readers with content that they are excited about and want to read. Real human insights are really important here. No one wants to read a boring wall of text that sounds just “okay.”

Most outlets have tools that detect AI-written content, so chances are that once it's flagged, no one will even take a look at what you've sent—no matter how much time you spent tailoring that prompt.

AI hallucinations are really common, actually. For example, we started to draft LinkedIn engagement posts for a client, and the AI started building an entire narrative about how creating multiple blog posts on a specific niche topic and choosing the right keywords resulted in a 40% increase in traffic and a boost in conversion rate by “x” amount. It was all written with exact numbers and in so much detail.

The catch – the client had only a landing page for their service and no blog posts at all on their website. But since these were “a common strategy” and “reasonable real-world results,” it was supposed to be a good story.

My advice is to remember that no matter how advanced it is, AI is still only a tool. It's here to help you, not to do the work instead of you."

AI misrepresentation will destroy the valuable trust with the media

Chad Walding, Chief Culture Officer & Co-Founder at NativePath

"AI created press releases and reactions do not get past the first cut often as they do not state values and contexts that are authentic. The AI content detectors are quite sophisticated now, although they may miss some, they can detect the patterns and structures of AI discourse, which is considerably less likely to attach editors. AI text lacks the nuances that are often in a human voice, which is exactly the voice that journalists use to connect with their audiences.

One common issue in AI content is known as "hallucination" where AI generates fabricated information or incorrectly presents context. For instance, AI might create a press release that indicates a startup product was featured in a publication, as opposed to indicating the product was not. This misrepresentation will destroy the valuable trust with the media. The best practice to use AI is as a tool to grammar corrections and structure, but not the drive voice of the real human entrepreneur."

Journalists want real-world knowledge than AI can't supply

Mark Kember, Head of Content at onebite

"Journalists want insight – that comes from your experience, from what you personally are hearing and experiencing in the market. They want attitude, not platitudes.

They want to inform readers in the most effective way they can, and that includes finding material that is more personal, more pointed or further ahead in terms of real-world knowledge than AI can supply.

AI-generated text is based on the most common combinations of words that has come before. By picking the words most likely to appear next to each other, you can create sentences that are likely to be semantically relevant, grammatically correct and linked to a general understanding of the topic. They are also likely to regurgitate previous talking points that don’t sound like you, and don’t reflect your individual understanding or your company’s specific messaging.

So why would a journalist use your similar sounding text that appears to say the exact same thing as everyone else? And why would you be happy to provide material that was kind of linked to the subject, but did not give you a chance to make that argument effectively for how your company views the world? AI can help you create something, but it might not be the right content for you or for the journalist."

Journalists receive AI-generated press releases full of hallucinated false facts

Mohamed Abd Elkhalek, Marketing Executive at Above Apex

"AI-assisted press releases fail because AI-tools are advancing in sophistication and along with journalists developing AI detection methods that filter synthetic-written AI-generated press releases before they are filtered through.

Typically, journalists receive AI-generated press releases full of hallucinated false partnerships, misquoted/lying from the industry leader, while abjectly misrepresenting revenue numbers - next internationally - what is worse, is that entirely false press release written by AI could even assert the company had received real awards that your company never won!

Quite frankly, AI systems do not have real experiences. Nor did AI in their data pools create language specific to the narrative of the entrepreneurs vision, and if they display something that resembles a press release, it is going to sound like poorly written copy from a cheap template with the same artificial text or text that lacked any tone/personality, and if a journalist reads that pitch, they too will move on because even they recognize the artificialness.

Your best practice: create your narrative as imaginatively and authentically as possible from your real experiences and observations - using your authentic voice, information and context that only you (and possibly you entrepreneur-doctorate) can know and then let AI improve your writing or improve your sentence structure flow."

AI can’t genuinely respond to the news cycle

Dennis Shirshikov, Head of Growth and Engineering at Growth Limit

"The press doesn't need more content — they don’t — but they need more substantive content. AI-generated releases rarely say anything you can’t get in a minute of Googling — anything specific or personal or urgent — and journalists can sniff that out in seconds. Instead, they tend to throw back up generalizations or jargony platitudes that can come out quite nicely, if with no voice, no thesis, no real story.

I’ve also worked with startups that thought they might be able to take a shortcut by getting ChatGPT or a similar tool to cobble together their Series A press release. It had the tone of a mishmash of a college-admissions essay and a mission statement from a decade ago — no tension or insight, no reason a journalist would need to care. Worse, the founder’s vision was lost in layers of plastic optimism.

When you pitch a journalist, it’s not just information you’re offering them. You’re proposing a story, and one that has to fit with what’s going on in the world, and in the industry, and in the journalist’s beat. AI can generate some patterns, sure, but it can’t genuinely respond to the news cycle, or call back to a dinner you had with your cofounder when the big idea finally clicked. That’s the sort of thing journalists react to.

One client turned in an AI-written founder bio that said she had spoken at Web Summit and written for Forbes. Neither of those things were true — she’d been to Web Summit once and was listed in a Forbes roundup. But the AI tool, fed a cursory sketch and a craving for assertion signals, had no trouble connecting the dots.

Another pitch carried a fake quote about supply chain interruptions caused by “recent sanctions,” events that did not yet occur. It was trained on old economic data, and it manufactured associations that sounded as if they could be true, though they had been pulled from thin air. These are insidious but destructive hallucinations. If and when a journalist does discover one, the credibility of the entire pitch — and often the company that issued it — erodes.

Founders should write from the details of experience, the conflicts and inconsistencies, as well as the openings and callings. Reporters don’t want to write about “a startup doing something innovative” — we want to write about a person taking a stand and doing something interesting with it.

Record yourself talking. Transcribe it. Edit from there. That’s a much better start than the blank prompt to ChatGPT. THEN and ONLY THEN let AI proof it for grammar or structure. But keep the soul."

AI-generated press content isn't trusted

Brandon Aversano, Founder at Alloy

"AI-generated press content gets ignored because it isn’t trusted.

Editors use detection tools to flag machine-written pitches. Once flagged, they’re trashed without reading. These tools aren’t perfect, but they’re good enough to filter out robotic phrasing, fake timelines, and generic language. If your pitch reads like a template, it’s done before it starts.

I’ve seen AI drafts claim partnerships that don’t exist, events that never happened, and timelines that don’t check out. One pitch said the founder “launched their third startup after exiting a fintech company”—when no such exit occurred. Another included quotes in the founder’s “voice” about market conditions they’d never discussed. These hallucinations break trust and damage your credibility with the people you’re trying to reach.

Real stories win. Reporters look for evidence, not exaggeration. Use AI only to clean up your grammar. Write the story yourself. Start with what’s real, growth data, lessons learned, product traction, and early customer reactions. Speak plainly. Be specific. If your business is early, say so. If you’ve failed before, say what you did differently. Don’t claim momentum you haven’t earned.

I write my pitches. Then I use tools to make them tighter. That’s it. I once landed a feature with a six-sentence pitch and a clean Google Doc full of facts and links. It worked because it didn’t pretend.

Write like someone who has something to say. Because if you don’t, no one will listen."

Media pitches require human involvement

Caitlin Agnew-Francis, Commercial Sales Manager at Desky

"Even when it comes to press releases and interview responses, which are both AI-generated content, the press is seldom able to pick up the content as it lacks the reality and context that journalists seek. AI content detectors have become quite advanced in order to trigger the flagging of content that is too formulaic, should lack the personal touch, or even be out of synch with current trends. To illustrate, an AI could generate a press release that reads nicely but contains no actual insights as to why the startup or entrepreneur is special. Such pitches tend not to consideration of the needs of the journalist in terms of new insights or context.

Hallucinations are easy to generate by AI, as it can either make things up or distort information. I have personally viewed AI-written pitches that have contained out-of-date information about the company, or have had broad statements such as, “This business is disrupting the industry,” with no supporting facts. Such fluff does not create credibility with either journalists or readers.

Good practices? It requires human involvement. Grammar and structure can be assisted by AI, however, the voice of the entrepreneur should not be substituted with an artificial one. I suggest running your content through AI to clean it up but be sure to always incorporate your own spin to it to ensure the pitch sounds authentic and relatable. Use of personal story or real life example will always give a pitch a chance in the crowded media."

AI misses your lived experience

Sahil Kakkar, CEO / Founder at RankWatch

"Using AI to write full pitches often has unintended consequences largely because AI misses your lived experience. Most journalists are interested in the founder's why not a robotic bio. Furthermore AI detectors are able to tag writing that is formulaic or disconnected from current events. we have even noticed AI-suggested pitches reference tools that are not even available yet.

Instead share your real story what started your business, what you are solving, what this work feels like. Let AI polish the grammar but let your voice be your voice. A combination of clarity and truth will always get more of a bite than something artificially polished."

Don't get flagged as synthetic content

Vaibhav Kakkar, CEO at Digital Web Solutions

"AI tools can be a strong force but they are not reporters or storytellers. Journalists can quickly see when a pitch is AI generated. For example these pitches lack detail and specificity when it comes to vague accomplishments or other industry contexts. We once saw a release of a company that claimed a founder won an award they never even applied for.

These hallucinations happen because AI guesses to fill in content gaps. When it comes to getting published founders need to speak from lived experience. A good rule of thumb is to create your story in your own words and let AI help you edit for flow. You will be able to keep the authenticity of the piece and not be flagged as synthetic content."

AI tends to produce overly tidy but generic fluff

Mimi Nguyen, Founder at Cafely

"My biggest pet peeve with founders is over-reliance on AI-generated interview answers and press releases when going after media coverage. AI can certainly help with tone or structure, but reporters can easily identify content that sounds artificial or generic. Press people need depth, not fluff, and AI tends to produce overly tidy but generic fluff without substance or specificity.

I’ve personally reviewed pitches where the AI invented job titles, exaggerated achievements, or used placeholder anecdotes that didn’t exist. These hallucinations instantly erode trust. Plus, many newsrooms use AI detectors now, which flag overly “robotic” language before an editor even reads the submission.

So what to do instead? Lead with your real experience whether that's a win with a customer, a bitter lesson learned, or an unexpected turn, and then craft your story from that. You can still utilize AI to squeeze the grammar or do formatting adjustments, but the core of the pitch must always be from lived experience. That's what people respond to and that gets printed."

The core story must come from the founder, not AI

Firdaus Syazwani, Founder at Dollar Bureau

"Editors want authenticity, not automation. Journalists are constantly pitched by founders. What makes someone stand out is their real voice, quirks, emotion, and relevance to current events. AI-generated content often lacks that spark. I once experimented with a fully AI-generated founder story. It was grammatically flawless but sounded like a Wikipedia page. Not a single outlet responded until I rewrote it with real anecdotes from my digital nomad journey across Southeast Asia.

AI content detectors are catching up fast. Publications now use tools that detect machine-written copy. If a press release or interview smells too polished or "generic," it gets flagged or ignored. Editors have shared with me that they often toss out submissions that look like they came from ChatGPT, because chances are, they did.

Hallucinations are a real risk. AI doesn’t understand truth the way humans do. I’ve seen AI fabricate statistics, misattribute quotes, or invent non-existent features in a product just to sound “complete.” That’s a liability, not a shortcut.

Best practice? Let AI refine, not replace. I recommend startups use AI for what it’s great at – tightening grammar, improving tone, or suggesting structure. But the core story must come from the founder. Journalists want to hear why you built your business, what scares you, what surprised you. That can’t be faked."

#PRontheGO