Brand copywriting has never been easier to create. In the age of AI, full-blown comms strategies can be drafted in seconds. Tools that last year felt revolutionary are now standard-issue — slick, fluent, and increasingly undifferentiated. The result? A flood of content that’s structurally perfect, tonally inoffensive… and, in most cases, instantly forgettable.

In this thinky-thought piece, I’m going to explore why taking a voice-first approach to brand copy is still the best way to land messaging that really sticks.

And, I’m going to explore how some brands have been able to leverage AI to optimise and scale this core principle — while others have been left disadvantaged by a superficial reliance on Large Language Models (LLMs) as a creative driver.

⠀“Content.” There I said it.

You can spare me your blushes — we're all grown-ups here. Content is a word that we tend to hear more and more. And yet the more we hear it, the easier it is to become desensitised to what it quietly erases: that 'it' being a real voice — full of meaning, intrinsic worth, and quality.

Patrick H Willems, filmmaker and commentator, breaks down his grievances within his own field in his video essay Everything Is Content Now.

In it, he frames content as a term born of late-stage capitalism — a linguistic marker coined by movie execs and big tech firms who, in an age of on-demand streaming, have become increasingly focused on profit through disposability and commodification. Predictably, he argues, when content becomes king, it’s often at the expense of craft — artful cinema flattened into optimised runtimes, with identikit storytelling that shuns creative risk in favour of algorithmic safe bets.

⠀

I see the same pattern emerging in digital marketing. As the barriers to entry have collapsed, brands now have limitless access to a production line of formulaic brand messaging — courtesy of fast, fluent language models that deliver on command. The temptation is obvious: spin up a strapline, draft a manifesto, populate a website. Hell, let's execute an entire product launch as simply as 1, 2... Ctrl + C.

Annnnd… Publish!

Of course, that speed comes at a cost, which amounts to a tragic erosion of creative freewill and individual brand expression. Because, unless guided with care, these tools default to a predictive median — a set-menu with only one dessert option: Vanilla.

Vanilla. Vanilla. Vanilla. Don't get me wrong. There's nothing inherently wrong with vanilla. But what it can't do is stand out from the crowd — exactly because it is the crowd: a confectionary cliché in itself that has been gobbled up by Language Models (along with every other cliché of course) to inform their baseline understanding of a world full of content without meaning.

This patterned understanding has given rise to a certain known quantity of generic AI prose. Writing that sacrifices the essential quality that makes brands shine (or taste like something with a bit more personality… Tutti Frutti, Rum & Raisin, or, dare we dream, Cinnamon Churro Crunch).

Now, like never before, brand copywriters need to start bringing the flavour(s).

Yes, let's play a game. I'm going to crack my knuckles on writing some copy. This copy is going to be SO good that, when you read it, you'll be SO sold on what I'm selling you, you'll want to tell me to shut up while angrily waving dollar bills in my face. I'm writing it right now. This is it. You're reading it — and you're buying into it even as it occurs to you I'm really talking about nothing at all. Well, I'm sold anyway.                       

                                                                                                              — Me, just now 

Ok, that was a little mad, and a little ad-lib — but it does have an identifiable voice.

So, what happens when we ask Google's Gemini language model to 'tidy it up'? Let's find out:

Let's play a game. Hear my knuckles crack as I conjure copy so persuasive, it will have you shoving dollar bills in my face, begging me to be quiet. I'm crafting it right now. This is it. You're reading it – and even as the absurdity dawns on you, you're already sold.

                                                                                 — Google Gemini, also just now

   

⠀

And immediately we've lost something. It’s not just that Gemini failed to clock a blatantly inserted Futurama meme, or that it deemed my closing note of self-deprecation somehow beneath it. It’s that… all of it suddenly feels flat and hollow and, well, vanilla — a perfect snapshot of Gemini’s default tendencies. And so it is, without competent creative direction these models are left to their own devices — and don’t so much write, as auto-complete their way to mediocrity.

This is a linguistic answer to the lowest common denominator. This is the ‘AI slop’ journalists warn their children about. This is the grey goo of the coming Textpocalypse (more on that later).

This is content — and it is the antithesis of quality writing.

⠀

⠀

Quality, more often than not, goes hand in hand with uniqueness and individuality. It’s what gives something that intangible sense of specialness. But, with everyone using the same tools, trained on the same datasets, and fine-tuned with the same ‘safe’ strategies, writing tends to hit something of a creative dead-end where sameness isn’t just likely — it’s baked in.

And, when every brand sounds polished but none sound distinct, the real threat isn’t just over-saturation — it’s an encroaching, pervasive sense of sameness. Brand copy that feels simultaneously both over-rehearsed and undercooked.

To make matters worse, studies suggest that even humans aren’t immune to this creeping sameness. According to a 2023 study from Cornell University — “The Effect of AI Assistance on Writing Style and Originality” — writers who used co-writing tools like ChatGPT or GrammarlyGO produced more formulaic, less distinctive work, even when prompted to be creative.

It goes even deeper. Academic institutions are increasingly reporting shifts in student writing that echo LLM structures — templated logic, overuse of passive voice, and generic phrasing. The implication for these trends is kind of worrying: AI isn't just echoing the median — it's coaxing us into doing the very the same thing.

⠀

So, writers who are experimenting with AI, but who still value their craft (if not their very souls), need to be vigilant — and they need to ensure that, in leveraging these tools, they aren’t lazily conforming to an algorithmic mode of writing. Crucially, they need to ensure the quality of their work isn't voided or displaced by tell-tale tidy syntax, safe rhythm, and content that's simply content to be… er... content.

Resistance is never futile —and that's not a bad strapline actually. But, if this mindset of defiance sounds a little too spicy as a brand direction, there are lots of positive countermeasures writers can employ to push back. How about simply refocusing on what we actively enjoy about writing? Or asking what made us excited to begin in the first place? What passions do we secretly want to communicate to our readers? And what observations still make us pause, or laugh, or ache?

But for now, an honest answering of the questions above is surely a good place to start if we want to create writing that’s rooted in something lived, not lifted. I'm talking about language with our fingerprints all over it — and a sense of a world that’s come from somewhere deeper than a dataset.

⠀

Whether it’s by culture or by code, the rise of template language certainly is flattening the tone of whole industries. A growing body of academic research confirms that brand homogenisation is real (a phenomenon also that's extending beyond language into the brand visual identity as well.)

In an excellent piece for Muse by Clio, Scott Harkey argues that tech innovation has quietly usurped creative instinct. Brands have become “systematised,” filtered through frameworks and dashboards, their voices shaved down to fit neat CMS modules and safe customer journeys. The outcome? A sea of brands that all sound, well… friendly, optimistic, helpful, genuine.

Sound familiar?

These adjectives — meant to codify a brand's individuality through words — often end up reinforcing a singular, agreeable median. Ironically, even the tools designed to encourage voice have started to sound suspiciously similar.

In a world of templated tone and commodified communication, there’s a genuine strategic edge available to brands who are willing to break rank. In market landscapes that have grown tonally flat, these rising stars will be ones daring themselves to be different: curious, candid, unapologetic, self-aware — as an example.

They can do that by investing in an organisational culture that embraces self-analysis — and isn’t afraid to journal, question, or provoke. If it feels like navel-gazing, so be it. Just as long as it works towards building an honest, authenticate messaging framework and the distillation of a unique brand voice that registers with an audience.

This doesn’t mean sending your team on a seven-day silent retreat in some remote Cypriot eco-village (come on, we've all been there). More realistically, it just means creating space for writers to reflect, stretch, and experiment with everything from founder story workshops, word association games, or even a game of mad libs with a few beers.

The goal? To dig beneath the surface and bear something real to the marketplace — something that's undeniably yours and essentially human.

So far, I know — it might sound like I’m being a little harsh on our robot minions.

But let’s be clear: this isn’t anti-AI evangelism. It’s not about gatekeeping tools or mourning the analog past. It’s about recognising that while AI excels at helping us make, it still struggles to help us mean — unless we give it something meaningful to work with.

And that’s the catch. Without a well-defined messaging framework — the kind that nails your value proposition, distils your elevator pitch, and articulates your brand voice with precision — Large Language Models are left to operate on scraps; minimal context, vague prompts, shaky assumptions. And when AI is fed generic input, it won’t hesitate to spit out generic output.

But here is another opportunity: with proper guidance from people who understand technical writing, content strategy, and Tone of Voice — people who know how to brief and calibrate a language model — brands can get the best from AI. If they’ve already done the foundational work.

The brands already thriving in this space aren't the ones shouting the loudest. They’re the ones who already know exactly who they are — and who have the clarity, confidence, and creative fluency to teach a system how to express that identity. To this end, being fluent in how Language Models work should be part and parcel of every copywriter's skillset —which comes with tacit understanding that this tool should be used to augment their writing, not replace it.

It won’t write your beliefs for you (...God help you if you ask it to). It won’t intuit your tone or invent your difference. But once you’ve defined those things — really defined them — AI becomes a powerful force for amplification across messaging, formats, and channels. In short, AI might not be a muse — but it sure is a fantastic multiplier(x10)

If the AI evangelists/doomsayers (delete as appropriate) are to be believed, there may come a time when AI reaches a new inflection point — one where, equipped with more advanced sensory inputs and multi-modal systems, it begins to sense and interpret brand expression in richer, more intuitive ways.

As I've mentioned elsewhere, I don't believe machine-based AI will ever truly achieve anything close to the spiritual definition of consciousness. But as its simulation becomes ever more sophisticated, I can see a point where the gap between human authorship and machine-led brand articulation may start to narrow. (And I do wonder if that day may be coming faster than many writers and creatives are comfortable admitting).

But for now, brands still need to tap into something machines don’t possess: a sense of originality born from taste, intent, and a gut sense of what feels right.

Originality isn’t just some a creative flex — it’s a measurable advantage. Increasingly, it points to better performance, and therefore more dollar signs. And, let's face it, dollar signs do look pretty!

⠀

Lovely. Now let's look at some data:

⠀

4x more engagement for posts that feature a clear point of view or personal experience, compared to generic tips or listicles. (Source: LinkedIn)

2.5x faster growth for brands that centre their messaging around values and belief systems, rather than just features or pricing. (Source: Edelman Trust Barometer)

Higher inclusion in AI-generated summaries for brands that use distinctive language, own specific terminology, or regularly publish original frameworks.

⠀

These stats reveal a clear correlation between the quality of brand messaging and increased engagement, memorability, and long-term value. This also relates to AI and the fact that search itself is becoming increasingly AI-driven (because LLMs naturally surface material that sounds more human). In this context, original language becomes indexable currency — in the form of phrasing that’s not only distinctive, but designed to be searchable.

As we’ve discussed, once you’ve laid the groundwork for your brand tone, AI can help scale it. But some brands have discovered an added value that goes even further — by using AI as a writing co-pilot, they’re able to harness its recursive training on what performs well in search to augment original voice with an extra layer of algorithmic intelligence. That means using AI’s predictive capabilities to fine-tune the structure of your messaging so that hierarchy, semantic clarity, and keyword alignment are all optimised for discoverability — while preserving the soul of your brand voice.

Take Rocky Brands, for example. The footwear retailer leveraged AI-powered content tools to boost their organic search revenue by identifying high-value keywords and refining on-page elements — all while maintaining a consistent, on-brand tone. The result? A strategic fusion of voice and visibility, and a clear demonstration that AI can do more than scale — it can sharpen, structure, and strengthen what makes your brand distinct.

Approaches like this — where brand voice is sharpened by AI and tuned for discoverability — have also sparked a deeper unease. If language can be optimised, scaled, and gamed so effectively, what happens when everyone does it?

In The Atlantic, Matthew Kirschenbaum paints a sobering picture of a near future he calls the 'textpocalypse': an era where synthetic writing — devoid of any real expression — doesn’t just coexist with human prose, but overwhelms it. He describes a kind of “planetary spam event” — a web increasingly flooded by machine-written text, stripped of authorship, intent, or even truth.

If that future holds — Wordageddon, The Great Content Spill, The Spammening, whatever we end up calling it — then professional writers may end up being the last meaningful line of resistance.

Because if the internet does start to write itself — faster, flatter, and with less fidelity — it’ll fall to us to to hold the line and slow the churn by insisting on craft. 

In practice that means:

Creating original frameworks and naming systems that resist templating — things AI can’t easily replicate or remix.

Crafting narratives with emotional architecture — arcs, reversals, tension, pacing — that machines struggle to generate with coherence or resonance.

Designing language that isn’t just descriptive but persuasive — messaging that shifts perception or inspires action, not just fills a text box.

Embedding cultural, contextual, and strategic nuance — the kind of stuff that LLMs can’t intuit because it’s based on lived experience, not training data.

Challenging clichés in real time — asking “why this word?” when the AI would’ve said “why not?”

Protecting brand voice as a strategic asset, not a style preference — curating tone, cadence, and vocabulary with long-term distinction in mind.

Building brand ecosystems that centre real authorship — commissioning thought leadership, attributing writing to individuals, and showing the craft behind the copy.

So yes, we may be headed towards a noisier web and a search landscape increasingly shaped by what predictive algorithm thinks sounds plausible. But, let me tell you, I can make predictions of my own — and I'm predicting that is AI's essential inauthenticity that's going make real writing — distinctive, intentional, human — not obsolete, but a much more valued commodity.

⠀

Like it or not, AI is changing the way we produce and consume the written word — and it’s only just getting started.

A comprehensive analysis by Copyleaks revealed an 8,362% increase in AI-generated web content between November 2022 and March 2024. By the end of that window, AI-generated text was present on 1.57% of all web pages analysed — a startling statistic, considering how recently this wave began.

Fast forward to today, and the figures are even starker.

According to a 2025 study by academic Dirk H.R. Spennemann, at least 30% of all text on active web pages now originates from AI sources — with the actual number likely approaching 40%.

For those of a nervous, AI-averse disposition, that no doubt sounds scary as hell. But traditional brands without a strong copy foundation in place? They might want to start getting a little twitchy too.

Research published in the snappily titled Journal of Business Research suggests this shift is already having a chilling effect on certain brand initiatives — with previously ‘safe’ names now facing a high-stakes contest for attention, as AI-driven copycat brands flood the market with a chorus of near-identical voices.

For brand copywriters on the frontline, I’d argue this isn’t a crisis — rather it’s a call to arms, and a chance to recommit. To meet AI’s template polish and faux-motional soundbites head-on, and to raise the stakes with writing that’s got real guts. Or, as Charles Bukowski — cult poet, misanthrope, and champion of the brutally honest — put it: “writing with blood in it.”

More recent studies are already suggesting that this kind of full-throated pushback won’t be in vain — with growing evidence now pointing to the intrinsic economic value of author-led writing within the context of machine-learned artificiality. It would seem global audiences are buying into it.

Take the NP Digital Study on Content Performance (2024): A comprehensive analysis found that human-generated content outperformed AI-generated content in web traffic by over 5.4 times. Over a five-month period, human-written articles consistently attracted deeper engagement.

This is the only real endorsement a voice-first messaging strategy ever really needs: human readers prefer human writing — with all its felt, lived, and instinctively attuned touchpoints across subcultures, shifting trends, and evolving zeitgeists.

The brands that embrace this human-led approach will always have the final word — because these are the brand voices that are really going to stick.