ChatGPT Mentions Your Brand. But Is Its Take Credible?

Happy Wednesday! On tap for today:

Let’s get started!

AI search, generative engine optimization, “making your stuff appear in ChatGPT” — whatever you call it, it’s increasingly important. Getting your brand into an AI-generated answer is a new goal for us.

But there’s an even newer one — and it’s harder to fix: when you show up, does anyone believe what the AI says about you?

That's the central question addressed in The Credibility Paradox, a new report from global communications firm Burson. They analyzed thousands of reputation-related prompts across seven major AI platforms, covering 85 companies, eight reputation levers and three audience types. The result was more than 55,000 believability forecasts — and some findings that should make communicators rethink where they're putting their GEO energy.

I connected with Steve Rubel, Burson's EVP of Media Insights & Measurement, to dig into what the research actually means for comms teams. His core argument: visibility without believability is exposure without influence — and building the latter requires a different kind of discipline than most teams are currently practicing.

Let's jump in.

✨ The report draws a sharp line between "visibility" and "believability" in GEO. For communications leaders who've spent the last year focused on showing up in AI answers, what's the most important mindset shift they need to make?

💬 Steve's take: Visibility is binary. You are either in the answer or not. The next generation of GEO revolves around how audiences react to an answer and what actions they take. As we like to say, visibility without believability is like exposure without influence. Believability is multivariate — it depends on the person, their demographics, psychographics and more. So yes, visibility is essential. But it's really the impression the answer leaves that matters even more.

🔍 Thanks. I was very surprised to see that workplace reputation — something typically left to internal communications — had an impact on that believability lever with general audiences. What should CCOs actually do differently with their employer brand content as a result of this research?

💬 Steve's take: CCOs need to view workplace not just in a vacuum, but as a reputation fulcrum — a force multiplier.

Employees leave a trail of evidence online, both positive and negative, and that evidence can shape other levers. For example, employee review sites attach hard proof to the leadership lever, which is normally difficult for LLMs to parse and qualify, and for audiences to interpret on its own.

The takeaway for CCOs is to consider infusing employee storytelling across as many programs as possible, since it can bolster the evidence required to build LLM believability.

📉 Leadership credibility ranked last across every industry you studied. How should executive communications strategy adapt?

💬 Steve's take: Executive visibility programs don't need to be rebuilt — but they do need to evolve. Our research shows that leadership ranks last in believability across all audiences, not because executives lack credibility, but because AI struggles to substantiate abstract claims. Vision statements without evidence don't travel.

The fix goes back to our strategy of building evidence ecosystems. Executives need to show how their vision is coming to life in ways others can independently verify — and that's where third-party voices become essential, including earned media coverage, analyst validation and expert corroboration. Together, they make leadership claims believable, not just visible.

🧩 You distinguish between companies that face a "credibility burden" and those that face a "translation burden." How does a communicator diagnose which problem they have — and does the strategic fix look meaningfully different depending on which category they're in?

💬 Steve's take: The first step is to conduct a diagnostic GEO audit with the right set of prompts that will field meaningful answers, which can then be measured for believability across different audiences. This will determine whether the burdens are grounded in claims that lack proof — a credibility burden — or whether messages need to be simplified, which is a translation burden.

The fixes will vary depending on what the diagnostic finds. In the case of a credibility burden, it may require going to market with industry or trade associations to address systemic issues in the category. In the case of a translation burden, it may mean a rethink on core message points and facts used across all communications activities.

🏗️ You propose something called “the 3C model” — Consistent Narrative, Compounding Authority, Chorus of Content. It sounds right in theory, but most communications teams are already stretched thin. Where would you tell a resource-constrained team to start?

💬 Steve's take: We would counsel them that this is not a revolution, but an evolution. The call to action is to return to the fundamentals of good, fact-based communications.

In other words: create simple, crisp narratives that are free of marketing speak and cascade them across everything you control; engage any and all content creators that LLMs might view as authoritative, no matter how small or large; and listen to what the internet is saying so that you can potentially steer it toward consensus and alignment.

📅 Where do you think GEO is in two or three years — and what separates the teams that will lead from those that fall behind?

💬 Steve's take: GEO is not just interesting for communications or even for marketers — we're hearing it's become a C-level topic. Over the next two years it will further magnify the importance of reputation and continue to elevate our field. Given its measurability, the teams that apply a GEO lens across all communications activities and harness AI to do so will be well positioned to lead.

Why this matters: Most GEO conversations to date have focused on presence — getting into the answer. This research reframes the question entirely: presence is a starting point, not an outcome. For comms teams, that means the work doesn't end when the brand shows up in an AI response.

Your next steps: Read the full Burson report and pay particular attention to where your sector falls in the credibility rankings — it'll tell you whether you're dealing with a proof problem or a clarity problem, which points to very different fixes. Then open any chatbot and ask it a few reputation-related questions about your organization. What it says, and how confidently it says it, is a reasonable first signal of where you stand.

A year ago, I joined Gab Ferree’s Off The Record, a community where 500+ comms pros learn together and learn from each other. Today, just after the one-year anniversary of OTR, Gab is celebrating Comms Bestie Day, a moment to share appreciation for the other communications professionals we all rely on.

There are lots of ways to celebrate. Visit the official Comms Bestie Day microsite to make a graphic of you and your comms bestie and share it on LinkedIn. Join Gab’s LinkedIn Live chat with Meredith Klein of Meredith & The Media at 11:15 ET today.

And if you want to give Off The Record a try — you should! — Gab is offering a great deal to The Comms Stack readers: a one-week free trial. Click here to join, and I hope to see you there!

Here’s a tip: don’t ask ChatGPT or Claude to add a long string of numbers. I tried to do that a few weeks ago — we collected gifts from a bunch of other parents, and wanted to know how much the total amount came to, and it was easier to copy/paste the string into ChatGPT than into the calculator app. To double-check, I then pasted it into Claude. The result?

Two different answers, but both were very confident they gave me the right one. And as it turned out, neither did.

That, Ephraim Cohen of FleishmanHillard notes, is going to start becoming a problem: “Highly polished AI outputs invite cognitive surrender. People start accepting automated answers simply because they sound authoritative, not because anyone actually checked them.”

Why it matters: Executives and other spokespeople may start relying on chatbots instead of relying on subject matter experts or communications pros. We’ve seen this in the practice of law, with lawyers using ChatGPT-written briefs that cite to hallucinated case law. If you work in crisis or exec comms, there’s a risk that someone empowered to speak on behalf of the organization will make the same mistake.

Your next step: Having an AI use policy is critical here — and, unfortunately, you’ll have to make sure you enforce them. But that’s probably not good enough. When you build your workflows, include steps that don’t involve AI at all. Requiring a five-minute web search (with AI previews disabled) to validate every fact asserted will save you hours in the long-run, and more importantly, help protect your reputation.

My next Claude Code for Comms webinar is next Wednesday at 1pm ET — sign up here. We’ll be building a web app that gives you a gut check on how a draft communication will land, giving you feedback on audience reaction, tone, clarity, and more.

To make sure we get through the build in an hour, you’ll need to do a bit of set up. It’s easy to do, and if you get stuck, no worries — Claude itself can walk you through it. Here's what you'll need to participate:

Seats are limited, but attendance is free! Hope you’ll join me.

AI adoption is critical for communications teams — and I can help. Let’s work together to get your team moving — or moving faster.

📋 A prompt playbook. Need a kickstart? My communications-focused prompt playbook has more than 600 prompts to help you get started. Download it, free.

🎓 Self-paced learning. I’ve created six courses to get you up to speed on how AI impacts communications. Get started here.

📅 A strategy call. Whether you're trying to build systematic AI adoption across your team or focused on AI search visibility for your brand or organization, I’m here to help. Learn more and book an intro call.

You can save 1 hour every day with Catch: the most proactive, capable, and trustworthy AI agent for admin work. Operating like a human employee, it securely handles your scheduling, email, and travel via phone, WhatsApp, and iMessage.

Start now for free

Looking for a role at the intersection of communications and AI? Here are some opportunities to check out:

Was this forwarded to you by a friend or colleague? Want to get this in your inbox next week? Click here to subscribe!

Quick question: how can I help?

What workflows are you struggling with? Where does AI still feel mysterious or overwhelming? What has worked that you’d like to share with others?

I’m a builder, and I’d love to help you and the rest of The Comms Stack community find great new ways to use AI.

Reply and tell me.
I read every response.

Even a one-sentence reply helps. For example:
“I wish AI could help me with ______.”

Until next Wednesday,

Dan

Advertise | Subscribe