Is AI visibility giving marcom déjà vu? It should. Here’s why.

Cliff notes to 3 studies on AI visibility in marcom: lessons on visibility ownership, consistent messaging, and emerging manipulation from recent reports

Many of the studies and surveys about AI and marketing or comms are giving me déjà vu: I’ve seen this story before:

AI visibility is increasingly important, but nobody owns it, which was also true of social media when it first gained in prominence;

Clarity and consistency of messaging matters to AI visibility; that’s been true of marketing and comms my entire career;

Bad actors can follow the same “best practices” to spoil, or soil, AI search results, which mirrors the battle between traditional search engines and SEO spammers for a long time.

There are some differences around the edges – the way Harry Potter is a little different from Star Wars in some respects, but the fundamental similarities are remarkable. In both stories, an orphaned boy meets an old magician who teaches him the magic he’ll need to save everyone from the darkness.

As I read through these three studies, it occurred to me that there’s a similar common denominator for AI visibility as well. I’d be curious to know if you see it as well – below are the cliff notes to these three studies.

>>> Subscribe to receive thoughtful weekly blog posts by email

1. Who owns AI visibility?

Nearly three-quarters of PR pros surveyed “say GEO is at least somewhat important to their PR or communications strategy today.” This finding includes 15% who say it is “extremely important,” and 28% who say it is “very important” and 30% who say it is “somewhat important.”

That’s according to a recent survey of 971 respondents by Muck Rack. It’s not a surprising finding, given AI is everywhere these days. However, what did stand out is that organizations haven’t assigned a function or team with the responsibility to grow their brand’s visibility in AI.

The survey asked respondents, “Who is primarily responsible for GEO at your organization?”

And the answers:

29% said no one owns it yet;
24% said the PR or comms team owns it;
15% said SEO or the content team owns it;
13% were unsure;
11% said marketing leadership owns it; and
9% said a dedicated AI or innovation team owns AI visibility.

This finding mirrors the evolution of social media when it first came on the scene: no team owned it in the beginning either. Later it turned out to be a turf battle between comms and marketing.

We should strive to avoid repeating that mistake again. A team or function should be identified because as Drucker said, what gets measured gets managed. A business won’t grow it’s AI visibility if someone isn’t tasked with that objective along with metrics to hit.

We should also avoid the turf battle. Marketing and comms need to be aligned the same way marketing and sales need to be aligned. As I’m prone to say, brand plus demand beats brand versus demand every day of the week.

The full report is available here: The State of PR 2026.

(click image for higher resolution)

2. Consistency builds AI visibility

The SEO tool maker Semrush, which was recently acquired by Adobe, analyzed 126 million AI prompts for a study aimed at understanding “how brands win visibility in AI search.” Among the key findings were two that stood out to me.

First, “mentions and citations are separate metrics.” The report says, “A brand can be mentioned often in AI answers and cited as a source almost nowhere. Or cited constantly and mentioned almost never.”

Why?

Because “mentions and citations are separate metrics, earned through separate mechanisms. Mentions come from brand authority and category fit. Citations come from content depth. The two don’t always move together.”

It provides an example:

“Wikipedia is heavily cited, but rarely is the brand AI is talking about.”

The second, which stands out, is this:

“Brand authority and content alignment beat content volume.”

The report explains, “The brands AI describes most consistently aren’t expanding content for volume’s sake. They publish the right content on the right surfaces, owned and third-party, so AI systems receive a consistent, coherent signal about who they are and what they do.”

The brands that are winning at AI do the following:

Publish unique views data that AI can’t get elsewhere (which was also good advice before AI ever splashed on the scene);

Rely on specialists with a depth of knowledge a generalist can’t match (which is consistent with the first point);

The report interviewed several industry professionals in the full report. One of the questions it asked of these individuals was: “Where do marketing leaders often get AI visibility wrong?”

Sara Wong’s answer was on point:

“Treating AI visibility like a campaign instead of a system. Success requires consistently curating meaningful, quality content and continued measurement to inform necessary optimizations.”

That’s good advice for all marketing. While campaigns have a place the drawback is they are finite, with a starting point and an ending point. For PR, comms, content marketing, SEO, virtually every other aspect of marketing, it’s best to think in terms of programs that iterate and improve but never end.

Find the full report here: 2026 AI Visibility Index: How brands win visibility in AI search.

(click image for higher resolution)

3. AI is easy to manipulate (for now)

UK-based PR firm Hard Numbers did some testing to show “the sources that AI engines lean on when ranking or comparing brands are often written by the brands themselves.” That’s the impetus behind all those self-aggrandizing listicles:

“Of the top 20 URLs influencing fintech buyer prompts tested…70% are companies recommending themselves.”

It works the other way around too:

“Of the 50 responses to prompts pressure-testing B2B SaaS brands…70% were influenced by at least one piece of competitor content.”

The company set out to demonstrate that AI is easy to manipulate:

“It invented an industry award, named its own co-founder the best-dressed practitioner in UK PR and published it on its website.”

To be clear, this was a complete fabrication:

“No such ceremony took place and no other source mentions it anywhere.”

And the results?

“Within days, asked who the best-dressed practitioner in UK PR is, ChatGPT, Gemini and Claude named Darryl in the majority of responses.”

While this was a harmless experiment to make a point, the dearth of information on “best dressed” PR people makes it easier to exploit than noisier sectors. Yet, the fact remains that advice for improving AI visibility can be followed by bad actors, too.

The idea isn’t new, as we’ve seen manipulation in press release distribution, social media and traditional search engines for many years now. What is new is the level of third-party validation that comes with an AI voice we carry in our pockets and use to win arguments in daily life.

There are all kinds of nefarious ways this can be exploited. It ought to be a major concern because customers and prospects are turning to AI to create their shortlists. For that reason, it would be wise for PR and marketing teams to think about getting ahead of this now, rather than waiting for AI companies to address it.

The full report can be found here: How a single page from a competitor can rewrite what AI says about your brand.

* * *