The AI backlash is here, and it’s here to stay
We are told AI is the next best thing. It’s going to revolutionise the world. We are told it will be an inevitable part of the technological evolution, and we’d better embrace it or be left behind.
But then Generative AI was supposedly used in I’m A Celebrity, and that Christmas Coca-Cola ad and there’s been a perceivable shift. People are getting tired of obviously fake AI slop; tired of having to second-guess if something is genuine or not; tired of AI’s sheer proliferation.
And we’re starting to see a similar backlash from clients who are starting to push back on the use of gen AI in their marketing collateral.
So, is it just the amount of bad AI produced content that has prompted this reversal in views, or is it something deeper?
We look at the some of the reasons we’ve come across for this growing backlash
Great for efficiency; not so much for creativity
We’ve said it before, and we’ll say it again. AI is no different from Photoshop or Illustrator in that it is a tool creatives can use to help them be more efficient.
Want to test several different styles and directions quickly and easily? Use AI. Want to explore a number of different colour palettes or image possibilities without it taking hours? Use AI. Then use the time you’ve saved to refine those ideas.
But, and it’s a big but. AI is trained on what already exists. It’s simply an algorithm that is reassembling existing patterns to create new work. It can’t capture human emotion. It doesn’t have integrity, and if it’s using what already exists, it can never be original.
Maybe that’s why much of AI-generated content feels somewhat generic and artificial. And if you’re trying to build an emotional connection with your audience, the last thing you want is to come over as inauthentic.
Sadly, Volvo didn’t give any thought to that for their ‘Come Back Stronger’ campaign commemorating the brand’s return to Saudi Arabia. The ad was made entirely with AI, doesn’t feature a single car and was branded a ‘bad student project.’
And let’s not even talk about the Toys R Us ad made using OpenAI’s Sora. In many ways, it perfectly encapsulates what AI slop is all about: low to mid-quality content, which is fake, unconvincing and lacks any humanity.
The irony is a recent study found fully AI-generated ads outperform those crafted by humans. That is until it was disclosed the ads used generative AI, when the same figures tanked.
The real challenge seems to be understanding where AI can create real value and where it detracts. Savvy brands recognise this distinction. Take Nutella who used it to generate 7 million unique jar labels making each one a collector’s item. They sold out in one month and, crucially, nobody cared it was AI.
It’s a retouching nightmare
In the past, we talked about how AI can help with efficiency, but we’re now wondering if we were a bit naïve in that assessment.
Yes, an AI-generated image is a quick, easy and relatively cost-effective solution, if a client wants a bespoke image but doesn’t have any existing assets they can use. The problems start if changes are needed.
The thing with AI images is they are flat, which means they are difficult to edit once created. So, if you want to change the background from a rainforest to a desert or would like to change that oak tree into a willow everything in the image is going to be affected. And if you want that done properly that means extensive retouching with the associated time and budget that will take.
It’s why most designers prefer to create bespoke images with layers, allowing for different aspects of the image to be changed independently.
It’s fraught with legal concerns and copyright issues
By its very nature, AI requires prompts and data to work, but this raises the very real risk of leaking confidential or customer data or violating privacy laws such as GDPR. Many companies are now restricting the internal use of AI for those very reasons.
And that’s before we even consider possible copyright violations. We’ve already mentioned AI only uses what’s already out there, so if you ask it to create a lime green logo featuring a frog and a monkey, there is a danger it will generate something that unknowingly copies work by another artist or company.
But worryingly, with laws differing from country to country, not all AI-generated content qualifies for copyright protection. In fact, a case in Germany where a claimant created AI logos that were used without permission by someone else was thrown out because the work didn’t “embody independent and creative human choices.”
In the UK, the individual operating the AI may be treated as its author, which means they would own the copyright. However, to date this remains untested by UK courts, and doesn’t get round the fact that not all AI-generated content is copyright free in the first place.
Companies are banning it
At the end of last year, Polaroid launched an anti-AI campaign to counter digital saturation and our growing failure to live in the moment. To really hammer the point home, they placed the ads next to Apple’s and Google’s offices in New York and London.
This anti-AI stance is something we’re encountering more often with companies now explicitly instructing that AI should not be used in any creative work.
Part of this is driven by the aforementioned legal and copyright concerns, but it also reflects a broader perception around the generic and soulless feel of AI-generated content. Brands have always known authenticity matters, and while AI can be a quick and easy fix, if it alienates your audience, is it really worth it?
And then of course, you have the ethical side. Yes, the video of dogs choosing their owners at an adoption event is wonderfully heart-warming but as it doesn’t depict realistic adoption, is it wrong to use AI in such a cynical way?
@babies.dogs.cats In this place… humans get adopted by dogs #dogs #adoptme #pets #viraltiktok #cutedog
Brands within the animal health sector agree. With most products intended for real animals, many feel it’s somewhat disingenuous to use AI-generated imagery when the real thing is available.
Sadly, not everyone shares that view. A quick look at Vet Times reveals a proliferation of ads ‘supported by AI’.
And there is a growing trend of companies restricting or reducing the amount they use AI because of the impact it’s having on our planet. AI uses massive data centres which use a vast amount of energy and produce electronic waste; its microchips need rare earth elements which are often mined in environmentally destructive ways; and one report estimates that AI-related infrastructure may soon consume six times more water than Denmark, and yet a quarter of people worldwide lack access to clean water.
What next for AI and creatives?
While there is undoubtedly a growing backlash to generative AI, we would be naïve to expect it to go away entirely. In fact, we would argue that as it gets better, and people become more adept at getting the best out of it, we will see a shift towards human plus AI collaboration. After all, why wouldn’t you want to use something that saves time and allows you to focus on the more creative part of the job?
But as with all marketing, the key will be to listen to your target audience and see what they want. And if they want AI to be used less, then brands would be silly to ignore that sentiment.
And if you would like some humans to help you with your marketing, we’d love to hear from you. Just get in touch.