Using AI Is No Longer a Competitive Advantage
Only two or three years ago, putting “made with AI” underneath a commercial was often enough to attract attention.
In 2026, that is no longer the case.
Generative AI has moved from an experimental specialist tool into the regular advertising production stack. Image generation, video, voice, music, virtual models, generative editing and even agentic production systems are now being used by real brands and agencies.
The advertising industry itself has acknowledged that shift. In 2026, Cannes Lions introduced dedicated AI Craft subcategories designed specifically to reward work where AI is essential to the concept, execution or impact, rather than merely being another tool somewhere in the production pipeline.
That distinction matters.
The relevant question is no longer:
Did this campaign use AI?
It is:
Did AI make the idea better?
The following twelve campaigns offer remarkably different answers.
1. Svedka: When AI Fits the Brand's Own History
For the 2026 Super Bowl, vodka brand Svedka revived its former Fembot character. The commercial was produced primarily using generative AI and turned its synthetic appearance into part of the idea rather than something to hide.
Importantly, this was far from a quick prompting experiment. According to the Wall Street Journal, the team spent roughly four months digitally reconstructing the Fembot, which had been retired for more than a decade, and preparing its expressions and movements for AI-driven production.
That is precisely what makes the case interesting.
A robot that looks artificial does not create the same uncanny-valley problem as a synthetic human. It is supposed to feel artificial. At the same time, Svedka used AI to revive existing brand IP instead of simply applying a fashionable AI aesthetic to the brand.
What brands can learn
The best AI ideas often start with the brand, not the model.
If an established visual language, character or brand universe naturally supports synthetic aesthetics, AI can amplify brand identity rather than dilute it.
The opposite approach is far weaker:
“We want to use Veo. What campaign could we make with it?”
Technology is not a creative brief.
2. Claude: Perhaps the Best AI Advertising Does Not Use AI as a Gimmick
One of the most interesting AI advertising cases of the year came from Anthropic.
For the 2026 Super Bowl, Claude released two darkly comedic films built around the questions “Can I Get a Six Pack Quickly?” and “How Can I Communicate Better with My Mom?”
The idea: people ask an AI assistant for advice, only to have completely inappropriate commercial messages injected into the conversation.
Claude positioned itself as an alternative to ad-supported AI systems.
In June, the campaign won the Film Grand Prix at Cannes Lions 2026. The Cannes jury highlighted the contradiction at the heart of the idea: an advertisement about a place where advertising should not exist.
What makes the campaign remarkable is not its technology.
It is its strategy.
Anthropic turned a product decision into a creative positioning.
What brands can learn
AI advertising does not have to be AI-generated.
A company can credibly position itself as technologically progressive while still using traditional directing, performance, humour and filmmaking craft.
That is an increasingly important lesson in 2026.
As AI-generated content becomes ubiquitous, the technology itself stops being the story.
The brand needs to provide the story again.
3. Aerie: “No AI” Becomes a Brand Positioning of Its Own
While many brands are trying to demonstrate how much AI they can use, Aerie chose almost the opposite strategy in 2026.
The fashion and lingerie brand expanded its long-running “Aerie Real” commitment and pledged not to use AI-generated people or bodies in its campaign imagery.
A campaign starring Pamela Anderson turned that tension into the creative idea itself. In the film, Anderson tries to prompt an AI system into generating models that feel “natural” and “unique”, only for the synthetic results to fail at the very human quality Aerie wants to represent.
According to Aerie, its original anti-AI announcement became the brand's most popular Instagram post at the time, earning more than 40,000 likes. The company also reported double-digit growth in brand awareness, while Q4 2025 sales were 23% higher year over year. These figures do not establish direct causation, but they suggest that the positioning resonated strongly with the audience.
What brands can learn
Not every brand needs to show as much AI as possible.
In a content environment saturated with synthetic imagery, verifiable authenticity can become a premium signal of its own.
The strategic question is therefore not:
“How much AI can we use?”
It is:
“What role should authenticity play in our brand?”
4. Gucci Primavera: When Technology Conflicts With the Brand Promise
In February 2026, Gucci published a number of AI-generated visuals around Milan Fashion Week as part of its “Primavera” communication.
The imagery included surreal human figures, fashion scenes, cars and highly stylised visual worlds. Gucci explicitly labelled relevant posts “Created with AI”, while the broader campaign also incorporated traditional photography.
The response was still highly critical.
Some viewers described the visuals as cheap or “AI slop”. More interesting than the technical quality itself, however, was the question of whether scalable generative production aligns with the value proposition of a luxury house.
Luxury does not merely sell a finished product.
It sells craftsmanship, materiality, scarcity, cultural capital and the idea that considerable effort goes into things that are deliberately not infinitely scalable.
That is where the conflict emerges.
What brands can learn
AI can work visually while still being strategically wrong.
Production innovation and brand positioning are two different things.
A value retailer may comfortably communicate efficiency.
A luxury brand frequently sells the opposite: effort, exclusivity and craft.
Brands considering AI should therefore ask more than:
“Does this look good?”
They should also ask:
“What does the way this image was made say about the value of our brand?”
5. Google Project Genie: When the Product Becomes the Campaign
Google demonstrated a different form of AI marketing with Project Genie in 2026.
Rather than simply producing a film explaining new technology, Google allowed people to experience the technology themselves.
Project Genie enabled users to generate interactive digital worlds using text, images or drawings, refine them through natural-language instructions and then explore those worlds directly.
The project won the Digital Craft Grand Prix at Cannes Lions 2026.
For brands, this matters because the boundaries between product, experience and advertising are increasingly disappearing.
The strongest demonstration of generative technology is often not a film explaining it.
It is the experience itself.
What brands can learn
For products that need explanation:
Show, don't tell becomes even more important with AI.
An interactive generator, configurator or personalised experience can simultaneously function as a product demo, content engine, social mechanic and PR story.
For software, automotive, interiors, fashion, entertainment and product innovation, this opens campaigns in which audiences can actively participate.
AI is no longer merely the production tool.
It becomes the medium.
6. Genspark: Speed as the Real Superpower
Genspark also used the 2026 Super Bowl to connect AI directly with a cultural moment.
Its commercial starring Matthew Broderick focuses on the notoriously unproductive Monday after the game: why work when Genspark can handle presentations, spreadsheets and other tasks?
According to Adweek, the spot was developed under significant time pressure. The company had roughly five weeks before broadcast, AI was used for elements including the script, and core production work was reportedly completed within a single day.
The interesting point is not whether every production step used AI.
It is the time-to-market advantage.
Traditional production is powerful when sufficient time is available.
AI becomes particularly valuable when an idea emerges late, the cultural context changes quickly or large numbers of variations are required.
What brands can learn
The biggest commercial benefit of generative production is not necessarily:
cheaper.
Very often it is:
being able to decide later.
Campaigns can be developed closer to the cultural moment. Products can be integrated faster. Social responses no longer need to arrive three weeks after the conversation has already moved on.
For marketing teams, that speed can be more valuable than pure production savings.
7. Coca-Cola: 70,000 Generations Are Not a “One-Click” Workflow
Coca-Cola is one of the major brands experimenting most visibly with generative AI.
For its 2025 holiday campaign, the company once again created generative interpretations of its iconic “Holidays Are Coming” world. Coca-Cola itself described two campaign films as AI-driven reimaginings of the original 1995 commercial.
The production numbers are particularly revealing.
According to reporting by The Verge and the Wall Street Journal, approximately 100 people were involved in the campaign. Five AI specialists from Silverside alone reportedly generated and refined more than 70,000 video clips. Coca-Cola also said the campaign could be produced significantly faster than its previous traditional holiday productions.
Despite that effort, the campaign again faced criticism for visual inconsistencies and its synthetic aesthetic.
What brands can learn
The case destroys two myths at once.
Myth 1: AI production means writing one prompt and being done.
It does not. High-end generation can require tens of thousands of attempts, selection, retouching, compositing and extensive art direction.
Myth 2: More iterations automatically create better creativity.
They do not.
70,000 generations can solve technical problems.
They cannot repair a weak creative decision.
We explore the actual production process in more detail in our guide to AI video production in 2026.
8. McDonald's Netherlands: Seven Weeks of Work Can Still Look Like AI Slop
In December 2025, McDonald's Netherlands released a 45-second Christmas commercial produced almost entirely using generative AI.
Instead of the familiar warmth of holiday advertising, the film depicted escalating festive stress and ironically framed Christmas as “the most terrible time of the year”.
The combination of a cynical concept and visibly synthetic human characters was poorly received. Within days, the commercial was removed from the brand's official channel.
The production process makes the case even more instructive.
According to production company Sweetshop, up to ten AI and post-production specialists worked on the commercial for around seven weeks, producing thousands of takes.
That is quite different from “we quickly made something with AI”.
Yet that was effectively how much of the audience perceived it.
What brands can learn
Production effort is invisible to the audience.
Nobody awards bonus points because a problematic shot required 400 generations.
The audience only sees the result.
More importantly, AI amplified a strategic problem that may have existed regardless of the technology.
A Christmas commercial designed around human connection requires emotional credibility.
If the people in that film feel visibly artificial and emotionally distant, the production method is working against the communication objective.
9. Popeyes “Wrap Battle”: AI Works Especially Well When Speed Is Part of the Idea
In the summer of 2025, Popeyes and McDonald's found themselves in a small fast-food culture war.
McDonald's announced the return of its Snack Wrap shortly after Popeyes had launched new Chicken Wraps.
Popeyes responded with an AI-generated diss track and music video.
The project was created with AI filmmaker PJ Accetturo. Suno was used as part of the music workflow, while the visual scenes ultimately relied heavily on Google's Veo 3.
The crucial detail: the team had less than three days.
That is exactly where using AI becomes strategically compelling.
A traditionally produced high-gloss commercial released three weeks later might have looked technically better.
Culturally, it would have been irrelevant.
What brands can learn
Generative AI is particularly powerful for reactive marketing.
Sporting events, memes, competitor activity, product launches and cultural conversations often have extremely short attention windows.
Here AI changes more than production cost.
It changes which ideas can be executed before the moment disappears.
However, this requires an organisation capable of making fast decisions.
If an improvised AI commercial still needs six approval rounds and three board meetings, the production technology has accelerated while the organisation has learned nothing.
10. Kalshi: The $2,000 Commercial Where the Flaws Became Part of the Concept
Kalshi's commercial during the 2025 NBA Finals became one of the best-known early examples of AI-native television advertising.
The film deliberately featured absurd situations: an alien drinking beer, surreal everyday scenarios and visuals that existed somewhere between commercial, meme and fever dream.
The message:
The world's gone mad. Trade it.
According to director PJ Accetturo, roughly 300 to 400 Veo generations produced only 15 usable clips. Production took two to three days and was reportedly completed for around $2,000.
Here, the unpredictability of generative video actually helped.
Kalshi sells prediction markets covering events ranging from sport and weather to unusual cultural developments.
A chaotic world fits the product proposition.
What brands can learn
This is not evidence that every commercial should now cost $2,000.
It is evidence that production method and creative idea can occasionally fit each other perfectly.
Kalshi did not use AI despite its strangeness.
The strangeness was the campaign.
The same aesthetic could be disastrous for an insurance company, private bank or premium automotive brand.
Context matters.
11. H&M Digital Twins: The Most Interesting AI Model Case Is Really About Rights
In July 2025, H&M released its first campaign imagery featuring digital twins of real models.
The AI counterparts appeared against backdrops inspired by international fashion capitals while presenting denim looks. Importantly, H&M accompanied the campaign with behind-the-scenes material designed to make the collaboration between technology, models, photographers and the creative team visible.
That transparency is crucial.
Virtual models immediately raise questions:
Who owns the digital likeness?
Who is allowed to use it?
Does the model continue to be paid?
Can the digital twin be altered?
What happens to facial and body data once the collaboration ends?
When announcing the initiative, H&M had already described a model involving real talent in the creation of digital twins and addressing ownership of those digital representations.
What brands can learn
The hardest question around synthetic humans may not be technical at all.
It is:
Who owns the person?
For professional AI production, consent, usage rights, duration, territories, manipulation rights and compensation therefore become central parts of the workflow.
A technically flawless campaign can otherwise become a reputational problem very quickly.
Responsible AI is not a footnote after production.
It starts at casting.
12. PUMA x Monks: From Individual Prompts to Agentic Advertising Production
PUMA and Monks presented a case in 2025 that may be more important for the direction of 2026 than many visually spectacular AI commercials.
Rather than simply generating individual images or videos, Monks created an agentic production workflow.
AI systems were connected across different stages of production, including insights, strategy, concept development, creative execution and asset production.
The output included a fully AI-generated PUMA film. Monks described its demonstration workflow as capable of creating a 30-second spot in a fraction of conventional production time, using technology including NVIDIA and Runway systems.
This may be where the larger transformation lies.
Not:
AI generates images.
But:
AI connects production stages.
What brands can learn
The next efficiency gain will not come simply from an art director prompting faster.
It will come from workflows.
A brand system can contain style guides, approved product assets, audience information, formats, campaign objectives and legal rules.
Automated systems can then generate, validate, localise and prepare variations for different channels.
The human does not necessarily disappear.
The role increasingly shifts toward:
strategy, creative direction, selection and quality control.
What Successful AI Advertising Has in Common
Looking across these very different campaigns, several clear patterns emerge.
1. AI Needs a Reason to Be There
At Svedka, synthetic aesthetics make sense for a robot brand character.
At Kalshi, unpredictability fits the product.
At Popeyes, speed justifies the technology.
With Project Genie, AI is the experience itself.
All of these are stronger than:
“We wanted to make an AI commercial.”
2. Brand Fit Matters More Than Technical Quality
A perfectly generated image can still be wrong for the brand.
Gucci demonstrates how easily AI can clash with concepts such as craftsmanship, heritage or luxury.
Conversely, deliberately artificial aesthetics can be completely appropriate when they belong to the brand universe.
3. Speed Is Often More Valuable Than Cost Reduction
Popeyes and Kalshi demonstrate one of the biggest practical advantages of generative production:
It shortens the distance between idea and publication.
For social media, reactive marketing and performance content, this is enormously valuable.
4. Human Craft Does Not Disappear
Coca-Cola involved around 100 people.
McDonald's spent seven weeks producing its commercial.
Svedka invested months preparing a consistent character.
The idea that professional AI advertising consists of one employee and a prompt window does not survive contact with reality.
5. Transparency Is Becoming Part of Brand Strategy
H&M shows the production process.
Gucci labelled AI-generated posts.
Aerie turned its rejection of synthetic people into the campaign itself.
The question “Is this AI?” is increasingly becoming part of how audiences perceive brands.
AI Slop: Why Democratised Production Makes Great Ideas More Important Again
The most important consequence of this development may initially sound paradoxical.
The better AI becomes, the less valuable AI itself becomes as a differentiator.
When almost every marketing team has access to high-quality image and video generation, a technically impressive image alone carries less value.
Supply explodes.
More images.
More films.
More content.
More variations.
And inevitably, more mediocrity.
That is why the term AI slop has become so relevant. It does not simply describe AI-generated content. It describes content where producing the asset appears to have required more thought than deciding why that asset should exist in the first place.
Scarcity shifts somewhere else.
Production is no longer scarce.
Attention is scarce. Taste is scarce. Great ideas are scarce.
That is an important development for brands.
The future of strong AI advertising will probably not be about automating as much human creative work as possible.
It will be about removing production barriers where they previously prevented strong ideas from being realised.
Our 6 Rules for AI Advertising in 2026
The current cases suggest six highly practical rules:
- Idea before tool. Choose the production method after defining the concept.
- Check brand fit. Just because AI can create something does not mean it belongs to the brand.
- Use AI's real strengths. Speed, transformation, variation, impossible worlds and personalisation are more interesting than merely simulating traditional production.
- Control its weaknesses. Products, people, logos, typography and continuity require additional supervision.
- Plan for human craft. Direction, art direction, editing, sound and quality control remain critical.
- Design for transparency. Consent, rights, provenance and disclosure belong in the concept from the beginning.
That may sound less spectacular than “the future of advertising will be fully automated”.
It is considerably more useful to businesses.
Conclusion: The Best AI Advertising Does Not Feel Like an AI Demo
By 2026, generative AI has firmly entered professional advertising.
Cannes Lions now has dedicated AI Craft categories.
AI-generated commercials run during some of the world's largest television events.
Global fashion brands are experimenting with virtual people.
Agencies are building complete agentic production pipelines.
But that also means the era in which simply using AI automatically appeared innovative is coming to an end.
The strongest cases increasingly demonstrate the opposite.
At Svedka, AI strengthens an existing brand character.
At Popeyes, it provides speed.
At Kalshi, its visual absurdity becomes part of the creative language.
At Google, it becomes an interactive medium.
Claude, meanwhile, won one of advertising's most prestigious prizes with a campaign about AI whose strength lies in its concept, performance, humour and traditional filmmaking craft.
That may be the most important lesson for brands.
AI is not a creative concept.
It is a new production medium.
And, as with every medium, the tool does not determine whether great advertising is created.
The idea does.
At GOODARTSY, we therefore combine creative strategy, directing, traditional production and generative AI. Depending on the project, the result may be a fully generative film, a hybrid production workflow or deliberately traditional live-action production.
Technology does not decide how the work should be made.
The idea does.
