The AI Campaigns Takeover Is Already Happening

I'll be honest with you: a year ago, if you'd told me that Mango would launch an entire fashion campaign for their teen line without a single human model setting foot in a studio, I would have called it a gimmick. A tech experiment that would look like a video game cutscene from 2010 and die a quiet death in some marketing team's post-mortem deck.

Then Zalando went and generated 70% of their editorial campaign images with AI in Q4 2024. H&M announced they're creating digital twins of 30 real models who will own and profit from their AI likenesses. Sephora's Virtual Artist has facilitated over 200 million virtual shade try-ons, and according to the brand, customers who engage with the tool are three times more likely to complete a purchase. 

This isn't a tech experiment anymore; this is the playbook, and we're all going to have to reckon with what that means for creativity, authenticity, and who gets to participate in the future of brand building.

The thing about fashion and beauty marketing that most people outside the industry don't fully grasp is that they’re operating on impossible timelines that keep getting shorter. A TikTok trend goes viral on Tuesday morning, and by Thursday, if your brand doesn't have content aligned with that aesthetic, you've already missed the moment. 

Brat summer, mob wife energy, quiet luxury (which came back around to quiet luxury being over, and now it's the messy girl aesthetic). The cycle never stops spinning.

Traditional campaign production takes six to eight weeks. You're booking photographers, stylists, models, and locations. You're coordinating schedules across multiple time zones. You're waiting for post-production. By the time you've gone through all of that, the cultural moment you were trying to capture has already moved on.

Zalando cut that timeline to three to four days. Not weeks. Days. And they slashed campaign costs by 90%. That's not an optimization, that's a complete restructuring of how content gets made. Their VP of content solutions, Matthias Haase, put it perfectly when he said AI gives them the ability to move at the pace of culture, spotting a trend and producing tailored content in under 24 hours.

According to Zalando's corporate communications, they can now 'react to trends like brat summer in under 24 hours instead of waiting weeks.' Their AI-generated images illustrated recaps of the year's biggest trends, including brat summer, mob wife, and double denim.

That's the new competitive landscape. Speed isn't just an advantage; it's becoming table stakes.

The ethical questions around AI models have been simmering since Levi's got absolutely dragged in 2023 for announcing they'd use AI-generated models to "increase diversity" (the backlash was swift and deserved, and they paused the program). The fear has always been obvious: AI replaces human models, photographers lose work, creative jobs disappear into algorithms.

H&M is trying something genuinely different. They're partnering with actual models and their agencies to create digital replicas, and here's the twist: the models themselves own their digital twins. Not H&M. Not some faceless tech company. The humans.

Model Vilma Sjöberg described seeing her digital twin for the first time as both exciting and unsettling. "It's a picture of me, but it's not me," she said. That cognitive dissonance is probably how most of us feel watching this unfold. The technology is impressive. The implications are still being written.

Louise Lundquist, a business development manager at H&M, explained that compensation will work the same way it does now. If a model does an e-commerce shoot, they're paid for usage rights to their image. "This would be exactly the same," she said. "It's the digital twin being compensated for the usage rights of the digital twin." 

The models can even license their digital selves to other brands, including H&M's competitors. (H&M's first AI digital twin campaign featured their denim collection - H&M denim), which feels fitting given how foundational jeans are to fast fashion.

Will every company be this scrupulous? Lundquist herself acknowledged that's not guaranteed. At least someone is asking the right questions before deploying the technology, rather than apologizing after the fact.

Here's where things get genuinely complicated. Coca-Cola's AI-generated Christmas ad last year was supposed to be a triumphant moment, a tech-forward reimagining of their iconic 1995 "Holidays Are Coming" campaign featuring those beloved red trucks driving through snowy landscapes.

The internet response was brutal. "Soulless." "Devoid of any actual creativity." "Nothing like celebrating the spirit of Christmas with the most soulless commercial possible." One person joked it was the best Pepsi commercial they'd ever seen.

The fascinating part? When System1 tested the emotional responses to the ad through their research platform, consumers actually scored it at a perfect 5.9, the maximum rating. The measured sentiment was overwhelmingly positive. The vocal backlash from creatives, artists, and industry observers still dominated the conversation.

This split between how people react to AI content when they're just watching versus how they react when they know it's AI feels like the central tension we're all navigating. The ads can be technically effective at driving recall and purchase intent while simultaneously making people feel like something essential is missing.

Coca-Cola doubled down this year, releasing updated AI-generated holiday ads with improved craftsmanship. Their global VP and head of generative AI, Pratik Thakar, told The Hollywood Reporter: "Last year we decided to go all in, and it worked out well for us. Consumer engagement was very high. Yes, some parts of the industry were not pleased, but that's part and parcel of doing something pioneering."

Is that confidence or cognitive dissonance? Maybe both.

Not every AI integration is about replacing human creativity. Some of the most interesting applications are actually expanding what creators can do.

Nike's recent Travis Scott collaboration used over 5,000 images generated through Midjourney during pre-production to craft the visual identity. The creative director Jak Bannon was transparent about it, posting the process on X and showing how AI-generated concepts were then brought to life through traditional VFX and production techniques. "AI is a marvelous tool for exploring ideas and digging deeper into what's possible," he wrote. The keyword there is "tool," not "replacement."

Prada went even further by using AI to actually create a fragrance ingredient. Their Paradoxe Virtual Flower perfume features what they call the "Jasmine AI Accord," developed by feeding an AI model data on hundreds of variations of floral species. The AI helped perfumers explore new possibilities and compose unexpected accords. That's not replacing the perfumer's art; it's giving them tools that didn't exist before.

Sephora's Virtual Artist solves a genuine customer pain point. The AI-powered try-on tool uses augmented reality to let customers virtually test makeup before purchasing, reducing the anxiety of buying products like foundation or lipstick that might not match your skin tone. 

The results speak for themselves: a 35% increase in online conversions for items tried via AR and a 30% reduction in returns. That's AI serving the customer experience, not just cutting production costs.

McKinsey dropped a statistic that should stop every marketer in their tracks: over half of consumers say they don't trust beauty content created by generative AI. At the same time, influencer trust is declining too, with consumers looking to influencers for beauty ideas dropping from 33% in 2023 to 25% in 2025.

So consumers trust AI less AND they trust influencers less. Where does that leave brands trying to build authentic connections?

Maybe the answer is transparency. Mango includes disclaimers on their website when images are AI-generated. H&M commits to watermarking AI-created content. In the opposite direction, Aerie launched a "100% Aerie Real" campaign that explicitly commits to never using AI-generated images, building on their existing promise to never retouch models.

Both approaches can work. What probably won't work is pretending you're not using AI when you are, or using it in ways that feel manipulative rather than useful.

After diving deep into all of this, I’m sure AI in advertising isn't going away. The economics are too compelling, the speed advantages too significant, and the technology is improving faster than the cultural resistance can adapt.

The brands that win will be the ones who understand that AI is incredible at efficiency and terrible at soul. It can generate content faster than any human team. It can personalize at scale. It can respond to cultural moments in hours instead of months. What it cannot do is replace the strategic vision that decides what's worth saying, the creative intuition that knows when something feels right, or the human empathy that understands what audiences actually need.

We're all going to have to figure out where to draw those lines, probably while the technology keeps moving and the lines keep shifting.

The question isn't whether to use AI. The question is how to use it in ways that amplify human creativity rather than hollowing it out, that respect the workers whose livelihoods depend on this industry, and that maintain the authentic connection that makes people actually care about brands in the first place.

What's your take? Are you seeing AI change how your brand approaches content? I know the conversation is worth having. 

Have ideas for future newsletters?
Hit reply and let me know what you'd like to see!

XX

P.S. Did you love this edit? Take 5 seconds to forward this to a friend who could use some no-bull advice and insights to get them through those “life is life-ing” days.