The Three AI Gaps Comms Leaders Need to Close

Happy Wednesday! This week:

Let’s get started!

Over the past few weeks, I’ve had a few conversations with communications leaders, all asking me the same thing — how do I get my team to be more AI forward? I’ve also spoken with a bunch of practitioners — some consultants, some at agencies, some in-house — and the theme is the same: what do I need to do to get the most out of AI?

I’ve been giving consistent advice. Outline why you’re using AI, not just what you’re doing with it. Direct your learning and training efforts toward that purpose. Establish guardrails for yourself and your team to account for AI hallucinations and our understandable mistrust of the machines. And don’t be a user of AI tools — use AI to build the tools, workflows, and solutions you and your team need. (And because that’s a lot to tackle, bring in someone to lead it.)

A new report from Axios HQ and Off The Record puts hard numbers on a lot of this: AI adoption in communications has moved well past the "are we doing this?" phase.

We’re using AI — that’s great. The question now is whether you're using it in a way that actually changes how your function operates — or just makes your drafts faster. And unfortunately, it hasn’t had the change we’re all hoping for. The Axios and OTR report highlights three of the gaps I hit on above.

First, there’s training and then there is training. Bringing in an outside person to talk about prompting — sure, that’s something. But as the report shows, we have to move past that point.

The single biggest predictor of AI maturity in comms isn't the tools your team has access to — it's whether anyone taught them how to use those tools for actual communications work. Among the advanced cohort — the 42% of respondents who build workflows, use custom GPTs, or the like — 58% work in organizations where AI training has been deployed, compared to 43% of non-advanced teams. The gap looks small until you see what it produces: advanced teams are more confident, more likely to build, and more aware of the risks around them. Training isn't a nice-to-have. It's the on-ramp to everything else.

So what does good training look like? It’s not a one-off webinar. If you don’t have someone dedicated on staff, it’s a months-long engagement with someone like me, building the manifesto, and then working hand-in-glove with your team to build those workflows or custom GPTs or the like. If you’re a solo practitioner, getting a part-time coach — an hour or two a week — can help a ton. It’s not that you or your team can’t do it on their own. It's hard to figure out where to start, to purposefully set aside the time to do it, and to get past hurdles and reinforce the wins on your own.

Comms teams are here to put out the fires, not start them. We can’t be producing work that has incorrect information, violates brand guidelines, or … worse. AI use puts those fears to the test. Almost half of the comms leaders polled — 43% — cited distrust of AI output as the single biggest barrier to broader adoption. A similar amount — 45% — have already caught AI-generated content that could have put their brand at risk. These are problems that we simply can’t risk running into. The Axios HQ/OTR report doesn’t have longitudinal data, but I’m willing to bet that had you asked the same questions six months ago, you’d have seen similar results. And without purposeful change, you’ll see similar results six months from now.

Yes, we have near-universal AI use. But we also have persistent distrust.

That tension doesn't resolve itself. It resolves when organizations build the guardrails that give practitioners permission to trust their own judgment about when AI output is good enough — and when it isn't. One in five organizations surveyed still has no guardrails at all. For a function where credibility is the product, that's not a governance gap. It's a liability.

I’ll go a bit further than the report here. We know guardrails are needed (well, 80% of us do!) but it’s also easy to pay lip service to that goal. “Nothing AI written can be released without a human review” is easy to comply with in fact but not in principle — it’s not effective governance. Each team (and each individual practitioner) will have to audit their own workflows to determine what truly keeps them safe from the types of mistakes that can cause major downstream problems.

At the same time, the point of AI use is to make us more efficient, thorough, and impactful, and the guardrails we place on AI use should fit that same goal. Guardrails need to be designed to fit your work, not the other way around.

This is the gap that will define the function over the next two years. Fifty-eight percent of comms leaders have built none of the structured AI assets — custom GPTs, automated workflows, etc. — that characterize advanced practice. Most teams are still at the drafting layer: faster first drafts, quicker research, cleaner edits. Or they’re using AI as a robust search engine. This is useful — but not transformational. The practitioners who are pulling ahead aren't just using AI differently. They're building things: media research tools, brand voice agents, monitoring tools that flag narrative shifts instead of just volume. That's the distance between AI as a personal productivity tool and AI as a function-level capability. Closing it requires someone with a communications background and the mandate — and the curiosity — to go deeper than the chatbot interface.

If you’ve joined my webinars, you already know I’m a big believer in the user-to-builder shift. But I want to caution you here. Get the basics down first, and understand the need for guardrails even if you don’t have governance in place. Don’t try to make complicated solutions (which aren’t actually complicated, as you’ll see once you start) until you have a solid foundation — you’ll get frustrated and give up as mistakes pile up. Becoming a builder is your goal, not your starting point.

Why the Axios HQ/OTR report matters: It’s confirmation of what we already knew and feared — AI maturity is now the differentiator, not adoption. If your team doesn't have a training plan, a governance document, and someone explicitly tasked with moving you from using to building, you're not behind the curve — you're defining it for yourself by default.

Your next steps: Depends on where you are with your AI use.

Posting everywhere means rewriting one idea six times, so you post to one, or none. SureThing turns one idea into native posts for every platform.

Put It on Autopilot

Press releases are narratives — sentences, paragraphs, full thoughts. But maybe they shouldn’t be. Maybe they need to be bullet lists.

Maybe.

The University of Texas MD Anderson Cancer Center had a significant research finding to share last fall. Their comms team didn’t write a traditional press release — they opened with three bullet-point takeaways, moved into a tightly organized facts section with an early prominent quote, and closed with question-led headings and detailed answers designed to surface in AI-generated search responses. (Check out the release here.) They were writing for the LLMs, not the reporters.

The release didn’t result in immediate visibility on ChatGPT or Claude, although the Center is optimistic that those results will come in time. But surprisingly, the format didn’t hurt them with reporters — and it may have even helped. As PR Daily reports, the Center landed coverage in nearly every major outlet including Reuters, the AP, and The Washington Post, more than 1,500 total media placements, and over 54,000 visits to the institution's newsroom website. The tradeoff — AI-optimized, perhaps at the expense of traditional journalists — never happened. The reporters seemed to like it, too.

That may have been the plan from the start. The PR director credited the format with helping journalists better understand a complex topic — it’s not just LLMs that like information structured in a clear, digestible way. This bears more study, of course, but the initial findings are massive if they turn out to persist. Maybe you should be bold — the format that works for AI turns out to be the format that works for humans under deadline. Go with it.

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

If you're trying to build systematic AI use across your team or just need a push getting started, I’m here to help. Learn more and book an intro call.

On July 16th at 1PM ET, beehiiv is going live with a look at the future of publishing, audience growth, and digital business.

What started as a newsletter platform has evolved into something much bigger: a place where creators and brands can grow, monetize, and own their audiences without stitching together half the internet to make it work.

The next chapter starts live at the Summer Release Event. 

Join us to see what’s coming next.

RSVP now.

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