The Identity Crisis: Is AI making every brand look the same? 

Dr Daniel Hulme
Chief AI Officer
WPP 

There is, of course, a legitimate fear that if everyone is using the same models they will end up making the same work. And some will, if they treat the tool like a vending machine and ship whatever drops out. 

AI can spin up a thousand renditions of an idea in a minute; what it can't do is tell you which one is any good, or what is really, truly original.

I call this the Monet Problem. Ask a model for a campaign "in the style of" anything and it returns endless plausible versions, but perhaps none that are truly original. Picking the one worth taking forward or building on requires discernment and it always did. What AI does is infinitely scale the number of options we can choose from. And, faced with too many similar options, people often freeze - precisely the problem branding was invented to solve.  

This is what we need to plan for - managing AI's inputs and outputs and making good decisions. And a key part of that is truly knowing your brand and what it stands for. Not by generating the five nebulous words that everyone else uses (integrity, innovation etc.), but truly understanding the value that you bring into the world. 

Yes, there's a danger of AI eating itself: models trained on the average, generating more average, until everything merges into a beige middle-ground. But that only happens if we let the machine both generate and judge - and if we're not able to truly articulate what's different about our brand in the first place. Point it at a real view of the world and our place in it and it becomes a bigger box of matches - allowing you to test and iterate with unusual combinations, faster, and, through synthetic audience 'brains' like those we've built at WPP, giving you a read on how an audience might respond before you decide.  

Sara Chapman
Executive Experience Strategy Director
adam&eve\TBWA   

When you repeatedly train an AI on data made by AI, the interesting outliers and ‘weird’ bits disappear over time. Academics call this ‘model collapse’; a descent to the statistical average or the most frequently occurring answer.   

Rely on AI to create the wrong parts of your ecosystem and you get the same thing: creative collapse. A state where a brand world regresses to blandness that consumers simply scroll past. 

The most distinctive ideas are unpredictable, perhaps even a bit messy; it’s often that ‘weird’ bit AI discards that’s the start of a great idea. Humans recognise that bit of ‘grit’ in the story is what pulls at your attention and sparks feelings you’ll remember. That’s the essence of distinctiveness. 

But resisting creative collapse isn’t about saying no to AI; it’s about finding the right blend between human and machine. Every client we work with gets a human-made idea rooted in insight and emotion. But AI empowers us to understand consumer and cultural intelligence at a scale no human team could match.  

For example, our proprietary cultural intelligence engine forecasts where cultural shifts and consumer spend intersect so we build strategies that will outlive cultural whims. Similarly, using real consumer data to create synthetic representations of consumers helps teams deeply understand nuance across geographies and cultures. Blended with human-led strategy and ideas, this creates work that’s deeply relevant and impossible to ignore. 

Brandon Kaplan
COO
McCann New York   

AI doesn't erode distinctiveness - accepting its first answer does.

All models are built to be agreeable, and they default to giving you the answer, usually well-formed, smooth and confident. If you treat that as a finished thought, your work converges on the statistical average of everything that has come before. The thinking will be polished, on-brief and quite forgettable.  

At McCann, we codified this into our AI Constitution, which governs both our people and our agentic systems. Article XIII, ‘Keep the Friction,’ requires the first answer to be treated as a draft, never the solution. Our agents are trained to challenge lazy thinking, surface what we missed, and offer the non-obvious route(s) to be explored more. If we use AI to intentionally slow us down, challenge everything, we will ultimately get to very good and very human ideas. 

Plenty of AI tools deliver speed and efficiency, high-quality outputs with very little input, and every agency has access to the same ones. The real differentiator is an agency's approach to AI enablement. Ours is grounded in use cases defined by our department leads (the craft experts, not the AI ones) and, just as importantly, in the training. Every person is learning the McCann way to use AI, so the ‘how’ is embedded within our people, not the tools.  

Ultimately, the agencies that win won't be the ones with the best tool stack, but the ones that incorporate the tools best into the way they work. 

Tyler Berry
Creative Partner and Co-Founder
YeahNice 

The irony is that

AI-generated work, sold to brands as limitless creative possibility, often all looks like it came from the same studio.

It’s in the details - the soft light, the gloss, the fingers - but it’s also the overall vibe that comes from a strange lack of human decision-making. 

Some agencies are famous for a house style and that’s great when it sets you apart and becomes a calling card. But it’s a huge problem when the world starts seeing every new brand campaign lifted from the same set of components.  

AI never combines separate ideas in a novel way. It’ll never surprise you with inspiration from an unlikely source or go against the grain in a considered way. And it makes the average, acceptable ‘industry house style’ route dangerously easy. 

For an industry that’s here to express what’s unique about every business, this should be damning. It produces a sameness that audiences are already wary of. They can feel when something looks uncrafted and unsettlingly familiar. 

Most markets are fiercely competitive and a distinctive brand is sometimes the only edge a business has. Trading that for a perceived short-term efficiency will only make competitors with crafted creative stand out more. 

Freddie Campbell
Client Services Director, Creative, Content & Production
House of Communication UK   

Every agency now has access to the same models. Access was never the differentiator, and it still isn't. 

AI adoption doesn't erode distinctiveness on its own. It removes the barriers of cost, size and time that used to separate who could compete, which means the gap between a distinctive brand and a generic one now comes down almost entirely to the judgement of the people deciding what to keep, what to cut, and what to push further.

That judgement now matters more, not less. AI will happily generate a hundred competent, on-brief, forgettable options. It won't tell you which one is actually right for that brand, in that category, at that moment. That's still a human call, and it's the one call that determines whether the work stands out or blends in. 

Our approach has been to leverage the efficiencies AI offers to develop more territories and test more ideas, faster, backing the ones that would previously have felt too ambitious to risk.

A brand doesn't become more distinctive by generating more options, it becomes distinctive by picking the right one.

AI can widen the funnel. It can't make the decision for you. 

Ravi Pau
Head of AI Operations
Havas  

Yes, there is a genuine risk. AI has made it incredibly easy to create more content, faster, and that creates an obvious temptation for brands. The danger is that we end up with a sea of sameness: more content, more campaigns, all polished but somehow emotionally forgettable.  

Havas recently conducted research, which will be released in full later this year, looking at more than 21,000 AI-powered brand experiences across seven markets, including the UK. The tension was clear: while businesses talk about speed and scale, people experience that in moments, such as chatbots, content, product recommendations and more. Those interactions directly impact how they feel about a brand and when it’s not good…it’s not good. Nearly two-thirds (57%) of people think that brands are using AI to create shortcuts. This creates a real threat for brands fighting indifference.

AI is squeezing the heart out of good work, producing brand experiences that are technically fine but easy to ignore.  

The sharpest principle from our research was that task beats theatre. AI works hardest when it helps people do something: understand, compare or decide. We should be letting AI carry the load where the job is clear, and keep human judgement, strategy and craft where the brand must mean something. 

The brands that will pull ahead are those which are disciplined with their AI use. Not the ones that simply use it everywhere.    

Chris McKibben
CSO
Dentsu Creative 

I’m not worried about distinctiveness. In fact, letting a well-trained AI loose with robust brand guidelines might actually help. 

We should be worried about the human impacts. And I don’t just mean lay-offs - this is more insidious. Ours is a deeply human business. The difference between great and garbage has always been the unreasonable care, craft, and commitment of our best talent. Passion persuades. Obsession breaks new ground. The struggle is what makes it worthwhile. 

And AI can support. In an ideal world, it gives us all a promotion. We outsource the grunt work and focus on elevating what matters. 

But AI is also the most attractive invitation to laziness we’ve ever received.

In a business that always expects more, it’s too tempting. It robs us of the satisfaction of a job well done: an insight finally cracked, an idea slowly coming into focus. Choosing from an infinite set of options breeds ambivalence. When you haven’t earned it, you won’t fight for it. 

So my advice: become an AI ninja. Embrace its speed and prolificacy. Sell work with better visualisations. Stress-test your hypotheses. But use it to make you effective, not just efficient. Don’t let the ease of AI lull you into accepting good enough, in an industry where that has never been good enough. 

Nemanja Pantelic
AI Transformation Lead
VML   

AI can put brand distinctiveness at risk, but not simply because brands use the same tools. Problems start when brands feed them vague briefs and similar source material. When that's the case, it is hardly surprising when the output starts to look and sound alike.  

AI is very good at producing the most likely answer. But the most likely answer is hardly the best way to build a distinctive brand.  

The quality of the work depends on the thinking, source material and judgement applied along the way.  

Let’s not forget that AI scales the system it is applied to. That’s why, in our content automation work, we build brand difference into the system before generation starts: approved brand principles, product truth, tone of voice, terminology, strong examples and market context. Sometimes this also helps a brand understand what its tone of voice actually is.  

AI is useful for speeding up research, trying more directions, adapting work for different markets and handling repetitive tasks. What gives the work character still comes from people: what to say, what to leave out, which idea is worth pursuing and whether the result feels true to the brand.  

A final human review is not enough anymore. It cannot repair a generic starting point. Originality has to be present earlier in the process. 

AI is most useful when it creates room for better thinking, rather than replacing it.