Before you go crazy with the free AI, think about what it could cost your business.

Right now, businesses are being presented with an apparently irresistible proposition: why pay for photography, illustration or design when AI can generate something in seconds?

Type a prompt. Press a button. Job done.

Except your customers can press the same button, and they can tell when you have too.

That's where the apparent saving starts to look a little different.

If an AI-generated image makes your £50 million construction business look like a start-up with a Canva account, that's a cost. If your product imagery is technically wrong, that's a cost. If your communications start looking interchangeable with everyone else's, that's a cost.

And if the people you're trying to impress look at something and immediately think, that's AI, before they've thought anything about your business, you've spent some of your brand credibility to save a few quid on an image.

Because that's the trade-off we're seeing more and more businesses make without necessarily realising they're making it.

The cost of creating an image may be heading towards zero but the cost of creating the wrong impression isn't.

That's the bit we need to talk about.

People can tell. Even when they can't tell you why.

There's a slightly dangerous assumption around AI-generated creative at the moment: nobody will notice.

They do.

Research from Canva and Harris Poll in 2026 found that 70% of consumers say they can usually spot AI-generated ads because they feel like something is missing. More significantly, 78% said they would rather see advertising made by people, and 87% believed the best advertising still requires a human touch.

That's interesting to me as a designer because something is missing is actually a pretty good description of the problem.

It isn't always six fingers anymore.

The technology is moving quickly. The obvious giveaways - mangled hands, nonsensical lettering, strange faces - are becoming less.

What remains is harder to define.

It’s the overly polished lighting. The generic composition. The strangely perfect people. The visual clichés. The lack of texture. The typography that looks almost right until you actually try to read it. The mangled signage, invented labels and letterforms that no designer would knowingly let out the door. The sense that you’ve somehow seen this image before despite knowing that you haven’t.

In construction and manufacturing there is another problem: the details actually matter.

A generic audience might not notice that a piece of machinery couldn't work, a fixing is in the wrong place, PPE is being worn incorrectly or a building detail makes no engineering sense.

Your customers will. Specifiers, architects, engineers, contractors, installers and manufacturers spend their working lives looking at this stuff. They know what a site looks like. They know how materials behave. They know what their people wear and what their products do.

Artificial detail isn't a small aesthetic mistake when you're communicating to technically knowledgeable people.

It can undermine the very expertise you're trying to demonstrate.

Good design was never about knowing how to operate the software.

This is probably the biggest misunderstanding AI has exposed.

The value of a designer was never their ability to physically produce the thing. Knowing Photoshop doesn't make someone a designer any more than owning a saw makes someone a joiner.

The software is a tool. Design is the thinking that happens before, during and after you use it.

A designer is asking questions that a prompt doesn't necessarily know need asking.

Who is this for? What do they already know? What should they notice first? What are we trying to make them understand? What does the competition look like? What visual territory do we own? Is this unmistakably our brand? Is it technically credible? Does it work at six metres on an exhibition stand as well as six centimetres on a phone? What happens when this campaign becomes a brochure, animation, website, sales deck and vehicle livery?

And perhaps most importantly:

Is this actually a good idea?

That's judgement. It's built through years of looking, making, getting things wrong, learning why they were wrong and developing an instinct for what works.

AI can produce fifty options before I've finished my coffee. That doesn't mean any of them are the right option.

More creative doesn't necessarily mean more creativity.

This is where the economics of AI become interesting.

Adobe's 2026 research found that 76% of organisations report improvements in the volume and speed of content ideation and production through generative AI. That's a significant productivity gain.

But when everyone has access to similar tools, trained on vast pools of existing visual culture, increased production creates another problem:

Sameness.

Adobe's research among creative professionals found concerns around maintaining brand continuity, commercial safety and homogenised creative output. Its much larger 2026 creator study found that 42% of creators who say it has become harder to stand out specifically blame the growth of AI-generated content.

That's the paradox.

The technology gives us almost limitless ways to create, but used without direction, it can make everything look remarkably similar.

Professional marketing commentators are increasingly talking about exactly this problem. WARC has described an emerging “authenticity reset”, arguing that the explosion of cheap AI production risks filling the market with similar-looking material and making genuine distinctiveness more valuable.

Dentsu Creative's research reaches a similar conclusion from the other direction. 90% of CMOs said craft and design are vital to ensuring their brand remains distinctive in a world of AI-assisted production, while 89% said trust and taste will become more important.

That doesn't sound like the death of design to me.

The value is shifting from execution to judgement.

Some of the production work designers have traditionally done will undoubtedly become faster.

Good.

Nobody needs to preserve an inefficient process simply because that's how we've always done it.

AI can help us explore compositions, generate references, test directions, extend imagery, visualise ideas and accelerate parts of production. There will be applications we haven't even thought of yet.

But the easier execution becomes, the more important the decisions around it become.

Interestingly, the people using these tools most heavily appear to recognise that distinction themselves. Adobe's 2026 survey of more than 16,000 creators found that while 75% consider creative AI integrated or essential to their workflow, 85% believe the final creative decision should remain with them.

So, AI is brilliant at giving you possibilities, but someone still needs to know which possibility is worth pursuing.

Your brand isn't one graphic

There's another reason businesses need to be careful.

A brand is cumulative.

One questionable LinkedIn graphic probably won't destroy it, but brands aren't built - or damaged - by one thing.

They're built through repetition.

Every website page, presentation, tender document, social post, photograph, exhibition stand, video, brochure and sales document adds another little piece of evidence about who you are:

·      Professional or amateur.

·      Specialist or generic.

·      Careful or careless.

·      Premium or cheap.

·      Distinctive or forgettable.

Over time, those signals accumulate into an impression.

That's why designers talk so much about consistency. Not because we're obsessed with keeping logos 14mm from the edge of things, although admittedly some of us are.

·      Consistency creates recognition.

·      Recognition builds familiarity.

·      And familiarity helps build trust.

In B2B particularly, distinctive creative matters because customers aren't necessarily ready to buy when they see your marketing. Your visual identity has to help your business lodge somewhere in their memory until they are.

LinkedIn's B2B Institute has made the same point in its research into effective B2B creative: the strongest advertising doesn't simply put a logo on an idea; distinctive brand assets are integrated into the creative itself so the memory of the communication becomes connected to the brand.

That's considerably harder than generating a nice picture.

And no, this isn't an anti-AI argument.

It would be ridiculous for a design agency in 2026 to tell businesses not to use AI.

The genie isn't going back into the bottle, nor should it.

Kantar's research into AI-generated advertising is useful here because it cuts through both sides of the argument. They found that AI involvement itself doesn't determine whether advertising succeeds or fails. What matters is how well it is used.

When AI was integrated seamlessly, more than 40% of the ads Kantar examined landed in the top tier for branded cut-through. When AI use resulted in distracting or unnatural visuals, performance was poorer.

That's a far more useful conversation than AI good versus AI bad.

Use it. Experiment with it. Let designers push it. Use it to make work faster where faster is useful. Use it to make previously impossible ideas possible. Use it behind the scenes. Use it visibly when the idea demands it.

But don't use it simply because it's free, and don't confuse being able to generate something with having something worth showing people.

Free is not the same as cost-free

We're going to see an unimaginable amount of content produced over the next few years.

Most of it will be technically competent and a lot of it will look quite ‘nice’. And an enormous amount of it will be completely forgettable.

That's why I don't think AI reduces the need for designers. I think it raises the bar for them.

The designer's job becomes less about pushing pixels around and increasingly about taste, judgement, ideas, art direction, brand understanding, technical accuracy and knowing when the machine has produced something brilliant - and when it has produced beautifully rendered nonsense.

So by all means, go crazy with the free AI.

Just remember what you're putting at stake every time you press publish.

The image might have cost you nothing. Your brand has taken years to build.

Talk to us about using AI without losing the ideas, craft and distinctiveness that make your brand stand out.

Let’s create something worth putting your name on.