The Most Dangerous Word in Marketing

I want you to try something. Open your brand’s voice guidelines. Or your client’s. Or your competitor’s. Look at the tone descriptors. Warm. Approachable. A little self-deprecating. Conversational and casual but not sloppy. Lowercase-friendly. Maybe some emojis, but not too many. Human-sounding. Real.

Now open anyone else’s. Same industry, different industry. Doesn’t matter. You’ll find the same words in the same order. Everyone independently arrived at the same “authentic” voice. Which means nobody has one.

That’s the thing about “authentic”: it’s been used so often, by so many brands, to mean so many different things, that it now signals the opposite of what it intends. When everyone claims authenticity, the claim is noise. And possibly a tell.

This isn’t just a shift in semantics. At this point, AI can generate content that feels authentic (warm, vulnerable, specific-seeming), the word has become actively dangerous. It creates a false sense of security while doing nothing to establish real trust. And most of us haven’t caught up to that yet.

How we got here

For most of the last decade, “authentic” was a useful corrective. The corporate voice (the one that sounded like it had been approved by seven layers of legal and then run through a 1950s blender) was genuinely the problem. Brands sounded like robots. “Authentic” was the push back toward human. Fair enough.

So teams did the work. They loosened the tone, hired consultants, shot behind-the-scenes video. They went lowercase on social. They let personality leak through the seams. And it worked for a while, because the contrast was real. When everything else sounded like a press release, a brand that sounded like a person stood out.

But somewhere along the way, “authentic” stopped being a direction and became a destination. It went from “sound less corporate” to a box you check. A production value: something you apply to content, like color grading or a font choice. And the thing about production values is that everyone can copy them.

That’s how you end up with an entire market of brands that all sound warm, slightly self-deprecating, and lowercase-friendly. They all followed the same playbook to the same place and arrived at the same voice. The “authentic” voice became its own form of corporate normalization, just wearing a flannel shirt instead of a suit.

Then generative AI showed up

And now the whole thing breaks.

Because here’s what AI is extremely good at: feeling. Tone. Warmth. That sense of a real person behind the words. A decent model can produce conversational, vulnerable, “authentic”-sounding content in seconds, at scale, for almost nothing. The thing everyone spent a decade optimizing for (the feeling of realness) is now the cheapest thing in the market.

This is the part I keep coming back to: we built our entire idea of authenticity on how something feels and sounds, not on whether it can be checked, verified, or proven. “Authentic” describes a subjective experience. It’s not a property of the content. It’s a vibe. And vibes are exactly what AI is best at producing.

So when your entire trust signal is “we feel real,” what happens when everything feels real?

The answer, I think, is that the word becomes not just useless but actually dangerous. It gives you the sensation of having solved the trust problem while leaving you completely exposed.

I should tell you where this comes from

I’ve spent thirty years building the digital infrastructure that brands use to talk to people. Bespoke ad experiences, global creative platforms, service design. Event production and publishing systems for sustainability conferences. I’ve been inside the machine—not observing it from a strategy deck, but building the actual plumbing that carries the message.

And what I’ve watched, over and over, is the same pattern: something built with real craft and real human judgment gets absorbed by a larger system optimizing for scale. The thing that made it valuable (the specificity, the point of view, the grain of the people who made it) gets sanded off in the name of efficiency. I watched it happen to advertising (programmatic killed bespoke). I watched it happen to design culture (acquisitions killed community). And now I’m watching it happen to voice itself.

That pattern is what makes me think about authenticity as an engineering problem rather than a brand problem. But there’s a personal dimension too.

Early in my career, I found myself working for a “self-help guru” who’s USP was teaching “authenticity for leaders”. Talks about it. Executive coaching on it. A whole paid community built around the idea. And privately: none of it was real.

Coaching emails were paraphrased over voicemail and written by an assistant. Employees’ personal relationships were manipulated through gaslighting. People who couldn’t afford the program were convinced to pay for access to the “community” as “investments in themselves.” Once, an assistant called my doctor to reschedule my appointment that day because the timing was inconvenient. Then I was lied to about it. The performance was masterful. It worked for years, because everyone in the room genuinely felt like this was the real deal.

When I eventually left, I was badmouthed publicly at one of the conferences. A friend’s father was in the audience (not the kind of guy who generally makes a scene) and stood up and called it out for exactly what it was. In front of everyone. I wasn’t in the room, but I heard about it later. And it was the first time I realized the whole thing was a performance, and it had been for years.

Here’s what that experience taught me, and it took years to fully articulate: the problem wasn’t that the performance was fake. It was good. Really good. The problem was that “authentic” was a feeling being produced, not a structure being built. There was nothing underneath. No decision-making you could examine. No track record that held up to scrutiny. Just warmth, charisma, tough love and a voice that made you believe. Until you saw behind the curtain.

And once you did, it didn’t just destroy trust in one person. It destroyed trust in everyone who sounded the same. The whole category. That’s what happens when authenticity is a production value. One exposure poisons the signal for everyone using the same playbook.

RuPaul said it best: “We’re all born naked, and the rest is drag.” Every brand is performing. Every leader is performing. Every LinkedIn post is a performance. That’s not the problem. The problem is when the performance is all there is. I’d add: some people are just better at makeup and wigs than others. When you peel back the drag and there’s nothing underneath: no structure, no receipts, no history of decisions you can actually check.

So if performance is universal and “authentic” doesn’t work as a signal anymore, what does?

I’ve been thinking about this for a while now, and the pattern I keep seeing isn’t warmth. It’s not tone. It’s not vulnerability. It’s specificity. Not “authentic” specificity. Actual specificity. Details so particular to a specific person, place, or experience that they’d be genuinely weird to fabricate. The kind of detail that makes you think, “Nobody would make that up.”

After I left that job, I was really depressed. I felt like a failure, and I barely left my apartment for quite some time. When I finally started looking for a job again, I got frustrated with applications going nowhere (and this was before applicant tracking systems). So I submitted one with a haiku as the cover letter. I don’t even remember what the haiku said. But it got me my first role in digital marketing. Because in a stack of identical applications, one idiosyncratic, slightly unhinged gesture cut through. It wasn’t “authentic.” It was specific. It was evidence of a real person having a real moment.

That’s the distinction I keep coming back to. “Authentic” is a claim. Specificity is evidence.

And specificity is almost impossible to fake at scale. An AI model can produce one convincing specific detail. A good one, even. But it cannot maintain a consistent stream of genuinely specific, experience-rooted details over time without the seams showing. Specificity requires having actually been somewhere, done something, screwed something up in a way that left a mark. It requires the kind of lived experience that no amount of training data fully replicates.

Think about the voices you actually trust in your field. The people whose posts you stop scrolling for. I bet they’re not the ones with the warmest tone or the most polished vulnerability. They’re the ones who are particular. Their references are too specific to be generated. Their screw-up stories have details that no brand strategist would approve. They don’t sound “authentic.” They sound like themselves, which is a much harder thing to copy.

Why this makes brand people uncomfortable

Here’s where I expect pushback, and I want to be honest about it: if your trust strategy is built on tone and feeling (on sounding authentic), this argument says you’re building on sand. Not because you’re dishonest, not because your team doesn’t care, but because the signal you’re relying on is no longer distinguishable from noise.

The convergence problem is real. When every brand’s “authentic voice” sounds the same, and AI can produce that same voice for free, the signal carries zero information. You’re not cutting through anything. You’re adding to the pile.

The alternative isn’t louder or more persistent claims of authenticity. It’s making your voice unmistakable. Not through warmth-by-committee, but through detail so particular to your actual experience, your actual decisions, your actual history that it couldn’t have come from anywhere else.

Generic realness is gameable. Particular detail is not.

I realize this is a hard sell in organizations where “authentic” has been a strategic pillar for years. Entire agencies have succeeded simply by creating “authentic” communications. If your reaction is “but authenticity is about values, not words,” I’d actually love to hear that argument. Because I think the values version of authenticity is much closer to what I’m describing than the word itself suggests. The word is just getting in the way. It lets you feel like you’ve solved the problem by naming it, when the real work is structural.

So, when did you last see a brand describe itself as “authentic” and actually believe it?

Not feel it. Believe it. With evidence. With receipts.

And if something did make you believe, what was the specific detail that tipped it? Not the tone, not the vibe. The detail. The thing that made you think: only a real person, in a real situation, with real stakes, would say that. I’m collecting those examples and counter-examples because I think the pattern is the whole game right now. If you’ve got one, I’d genuinely like to hear it.