Most AI-generated brand messaging reads like marketing copy trained on even more marketing copy. But it doesn’t have to be like that. Here’s how the best marketing and creative teams are using AI to develop pin-sharp positioning and whip-smart brand lines.

Confidence gap

Many of us now use AI in some form. Blog content, social copy, summaries, captions and campaign variations can be created quickly, often with surprisingly competent results. So far, so good. But when the stakes rise – positioning, campaign messaging, brand voice or strategic narrative comes into play – confidence in AI tends to disappear. And we all know why. AI-generated messaging often sounds polished but generic. Strategically thin. The kind of copy that ticks boxes while saying very little. For organisations that care deeply about reputation and positioning (i.e. the global success stories), that’s a problem. Obviously.

The assumption is that AI simply can’t produce high-level brand language. Strategic messaging still belongs entirely to human writers. But increasingly, that assumption feels shaky. Because the real shift is not that AI is replacing writing. It is that writing itself is changing.

The problem with most AI prompts

Most teams are still using AI as an output machine. A prompt goes in, some copy comes back out and everyone hopes for the best. Which is usually how you end up with campaign lines that sound like they were assembled in a corporate escape room. The problem is not the technology itself. It’s the lack of context, direction and strategic thinking surrounding it. Today, a typical AI prompt for a campaign line or a piece of brand messaging might look something like this (and, to be fair, most of us are now using more detailed prompts, but let’s use the following for illustrative purposes):“Write a catchy slogan for a leadership programme. Make it inspiring, memorable and engaging. Keep it short and dynamic. See attached our brand voice guidelines for guidance on tone. Use wordplay if appropriate.”

Good luck with that.

Strong strategic messaging has never worked like this. Before AI, nobody expected a creative agency to develop distinctive positioning from a two-line brief and a vague adjective like “dynamic”. The same principle applies here. Better inputs create better thinking. The marketing teams getting real value from AI are feeding it considerably more context. Brand background. Audience insight. Internal tensions. Tone of voice examples. Competitive positioning. Personas. Even references to sentence rhythm, pacing and emotional tone. The strongest prompts increasingly resemble strategic creative briefs rather than instructions. Because that’s where things start getting genuinely good.

Writing using AI

The most effective teams are not using AI to write for them. They are writing using AI. And it’s an important point to note.

Using AI to write usually produces generic work because the model defaults towards probability. It selects the most statistically likely phrasing, which is why so much AI-generated marketing language converges around the same structures, clichés and predictable turns of phrase. The result is polished, technically competent copy that sounds just like everything else. Writing using AI is different. It becomes a collaborative process built around refinement, challenge and iteration. Ideas are tested. Routes are pushed further. Language is tightened. And messaging evolves through a conversation rather than appearing fully formed after a single prompt.

It’s worth bearing in mind that, the strongest results rarely emerge in the first couple of hours. They develop gradually as teams refine the thinking, sharpen the tone and articulate what feels right – and what doesn’t. That process requires more human judgement, not less.

Why feedback quality matters

One of the biggest shifts AI introduces to the writing process is the importance of articulate feedback. The organisations seeing the strongest messaging results are often the ones capable of explaining nuance clearly. Not: “This doesn’t work. Try again.” But: “The tone here feels abrupt and slightly curt. It needs to have pace, but with warmer, more approachable language.” Or: “The sentiment here is fine, but we need to add some flair to the language. Try some repetition. Use the power of three in a sentence to add impact and tone.” This kind of feedback gives the model something it can work with.

In practice, the process starts to feel far more like a creative conversation than a transaction. A campaign line will probably be tested, rejected, rewritten and refined 50 times. Messaging becomes something that’s shaped collaboratively in real time. Which means that strategic brand messaging still depends heavily on craft. AI can process language at warp speed, but it still struggles with judgement, subtlety and implication. The writer’s sensitivity to tone, nuance and audience psychology remains hugely important. (Arguably, more important than before.)

Why most brand voice guidelines fail in AI

Brand voice is where many organisations start running into problems. And, to be honest, we’re also yet to learn how to stop AI from going off-piste with brand voice, but a big part of the problem is caused by feeding conventional tone of voice guidelines into the LLM. Most existing tone of voice guidelines were written for humans. And human writers are good at interpreting nuance. Give an experienced copywriter a handful of brand traits –  confident, intelligent, approachable, authoritative – and they understand how to balance those qualities.

AI tends to interpret them much more literally/weirdly. The language often becomes an awkward blend of stated characteristics rather than a genuinely coherent voice. The organisations getting this right use AI-specific brand voice guidelines. (And the smartest organisations ask SIM7 to create this for them, by the way.) These will include detailed, extensive examples of the brand voice in action, which are far more helpful for an LLM.

It’s still about craft

There is a temptation to frame AI as either a creative revolution or a creative disaster. But the reality? It’s somewhere in the middle. AI is exceptionally good at accelerating parts of the writing process. It can help teams explore more routes, test more language and develop messaging frameworks more efficiently than before. But strategic brand messaging still requires judgement, restraint and clarity of thinking. Human thinking.  

The marketing teams getting the most value from AI are not using it to avoid the work. They are using it to go deeper into it. Which means staying ‘inside the conversation’, refining continuously, challenging the language harder. And giving much, much richer feedback.

Basically, they use AI to sharpen strategic thinking rather than shortcut it. And that’s probably the most important shift of all. In future, the advantage will not belong to the organisations producing the most content. It will belong to the organisations that know how to shape language with greater clarity, precision and intent.

Because strategic messaging was never just about output.

It was always about craft.