Does this look fake to you?

When even school photos are sprouting extra fingers, you know something’s off. Bea Widdup digs into the AI aesthetic that’s taking over our feeds – and how brands can stop looking like they all used the same prompt.

Last Tuesday in the Satsuma office, someone pulled up a school photo of their child. It looked normal at first – bright smile, clean collar, classic blue backdrop. Then someone squinted.
“Wait… are those extra fingers?”

Sure enough, the school photographer’s “AI touch-up” had given one poor kid seven fingers and a ghostly thumb. Cue the collective “ohhh” that only happens when your brain suddenly realises what’s off. Now you can’t unsee it.

And once you do see it, you start seeing it everywhere – in your feed, your inbox, the ads between your Instagram Reels. Flawless lighting. Skin so smooth it looks like it’s never met a pore. Hands that don’t quite look human.

If parents are spotting it in school photos, your customers are almost definitely spotting it in your brand posts.

It’s good enough to fool you – until it isn’t

AI imagery has become scarily convincing. At a scroll’s pace, it passes the test. But stare for a second longer, and the spell cracks. There’s something sterile about it.

The algorithmic aesthetic – that perfectly composed, poreless, painterly look – has become the default visual language of 2025.

And that’s the problem.

When every brand uses the same ‘sunlight through a café window, Gen-Z holding a latte aesthetic’, everyone’s grid starts to look like it came from the same prompt pack. You can’t say you’re all about the personal touch when your visuals are built from stock images and extra knuckles.

The sameness effect

So, why does it happen?

AI image models are trained to optimise what’s already popular. They chase trend references, symmetry, and smoothness. In human terms, they pick pretty over real.

That means your AI lifestyle shot of friends laughing in a co-working space will probably look identical to 10,000 others generated this week.

It’s not just dull, it’s a trust hit. People might not always know why they feel uneasy, but they do. Subconsciously, we spot the sameness. We stop believing, because we don’t know what’s real and what’s made up.

Spot-the-fake

Here’s a quick scroll test you can run on your own feed:

Zoom to 125% and scan in this order:

• Hands: Too many fingers? Weird nails? Limbs that don’t obey geometry?
• Eyes: Vacant gaze? Catchlights that don’t line up?
• Text: Blurry menus, warped t-shirts, unreadable signage.
• Edges: Melting jewellery, blobby ears, bent glasses.
• Light: Shadows going rogue, reflections that disobey physics.
• Props: Repeating patterns, cloned cups, mysterious floating objects.

Once you’ve seen the signs, they jump out like bad Photoshop from 2008. I was only a baby then by the way, but still…

When fake looks foolish

Take the now-infamous Willy Wonka Experience in Glasgow. The ads were pure AI Candyland – cascading jellybeans, golden arches, misty chocolate rivers. The actual event? A cold warehouse with two sad balloons and a single smoke machine. Cue mass refunds and a media meltdown.

Or we can recall one of the Trump campaign’s mistakes, when supporters filled Facebook with AI images of Trump surrounded by cheering black voters, trying to promote his inclusivity. When the fakes were exposed, the point they were trying to make collapsed entirely. Instead of connection, they proved disconnection.

The takeaway here is that people will always ask why you chose to fake it. If the real version doesn’t exist, maybe that’s the story that needs to be told.

Making AI work for you (not instead of you)

AI isn’t evil. It’s just easily overused. The trick is to use it as a storyboard, not a billboard.

Here’s how to keep your visual soul intact…

• Use AI for ideas, not ads.
• Build mood boards, shot lists, set designs – not final posts.
• Make a visual style guide first.
• Define your brands language, banned words, and texture pack.
• Mix your media. Blend real people and real places. A quick phone photo beats a fake smile.
• Add data visuals or bold graphics, and limit AI to abstracts or conceptual fillers.
• Disclose smartly. On posts where AI is used, note it – or at least add human alt text to ground it in reality. Because authenticity – the imperfect kind – scales far better than algorithmic pretty.

The final look

AI lifestyle imagery is now good enough to trick your eyes, but not your instincts. The more brands lean on it, the more they start to blur together in a glossy soup of sameness.

So next time you’re about to post that effortlessly candid shot that took no effort at all, ask yourself, does this look fake to you?

And if you’re not sure, we’ll tell you straight.