Most generative AI advertising examples you see online come from brands with seven-figure budgets and in-house AI teams. That is not useful if you are a growth marketer trying to figure out what to actually build this quarter. Nike, Coca-Cola, and Nutella can absorb the cost of a campaign that does not land. Most teams cannot.
Late 2025 and 2026 gave marketers a wider set of examples to study, some polished, some genuinely controversial. Coca-Cola ran its AI holiday campaign twice, once in 2024 to real backlash and again in 2025 with a visibly more refined result, which is itself a useful data point on how fast the underlying technology and audience tolerance are both moving.
This guide breaks down real generative AI marketing examples worth studying, what made each one work or fail, and how a leaner team can apply the same thinking without an enterprise budget.
Key Takeaways
- Every campaign that worked paired AI output with a clear, deliberate creative point of view, not just raw generation.
- Fully AI-generated creative performs best in fashion and performance contexts, and worst in nostalgic, emotionally loaded brand moments.
- Reactive speed, producing a finished ad within the same news cycle that inspired it, is a genuinely new capability AI unlocked.
- The biggest gap between big brand examples and everyday marketing teams is not access to AI tools; it is access to a director and editor who know how to shape the output.
What Is Generative AI Advertising?
Generative AI advertising uses artificial intelligence to create or modify advertising assets such as copy, images, video, audio, virtual presenters and campaign variations.
Unlike traditional advertising AI, which is commonly used for tasks such as audience targeting, bidding, forecasting and campaign optimization, generative AI directly contributes to the creative output of the advertisement.
Generative AI vs. Broader AI in Advertising
| AI Use in Advertising | What It Does | Example |
| Generative AI | Creates copy, images, video, audio and creative variations | AI-generated video campaign |
| Predictive AI | Identifies patterns and predicts likely outcomes | Predicting which audience may convert |
| Advertising automation | Automates campaign decisions and workflows | Automated bidding |
| Personalization | Adjusts messages or creative for different users | Dynamic product recommendations |
| Creative analytics | Identifies which hooks, visuals or messages perform | Comparing ad variations |
This article focuses primarily on generative AI advertising examples—campaigns where AI played a direct role in producing or adapting the advertising creative.
Marketing teams are using generative AI to:
- Create video and image ads without full traditional shoots
- Generate and test multiple headlines, hooks and creative variations
- Produce localized versions of the same campaign
- Create synthetic models, environments and product scenes
- Adapt one campaign concept across different audiences and platforms
- Respond quickly to cultural moments and trending conversations
The most effective generative AI ads still depend on human creative direction. AI can increase production speed and variation, but marketers still need to determine the campaign idea, brand message, audience, quality standard and final creative decisions.
How Is Generative AI Used in Advertising?
Generative AI can support several parts of advertising production, but the strongest use cases solve a specific creative or production problem rather than adding AI simply for novelty.
| Use Case | What Generative AI Does | Example From This Guide |
| AI video production | Generates scenes, characters, backgrounds or complete sequences | Coca-Cola, Toys “R” Us |
| Image production | Creates models, locations and campaign imagery without traditional shoots | Mango, H&M |
| Ad copy generation | Produces and tests messaging variations | JPMorgan Chase |
| Personalized creative | Generates multiple versions for products, users or markets | Nutella |
| Reactive advertising | Produces campaign creative quickly around cultural events | Popeyes |
| Synthetic talent | Creates or recreates people when traditional filming is difficult | Under Armour, H&M |
The advantage is not simply producing more content. Generative AI becomes most useful when it reduces production constraints, allows marketers to test more creative directions, or makes campaigns possible that would otherwise require significantly more time or budget.
8 Real World Generative AI Advertising Examples
Here are some of the best generative AI advertising examples that show how brands are using AI to improve creativity, speed, and campaign performance:
1. Nutella’s Seven Million Unique Jars
Nutella used AI to create over 7 million unique jar label designs, ensuring that no two jars looked the same. The campaign turned a regular product into a collectible, encouraging customers to share their unique jars on social media without any promotional push.
Nutella used an algorithmic design system to create more than seven million unique versions of its packaging, giving each jar its own visual identity.
Rather than using AI to simply produce another advertisement, Nutella applied computational creativity directly to the product. Every package became part of the campaign, turning a mass-produced item into something that felt individually designed.
How AI was used:
AI-assisted design generated millions of visual combinations while working within Nutella’s recognizable brand identity.
Why it worked:
The technology supported a simple creative idea consumers could immediately understand: every jar was different.
What marketers can learn:
Generative systems are most effective when they enable an idea that would be impractical to produce manually. The technology should make the concept possible rather than become the concept itself.
2. Coca-Cola’s AI Holiday Campaigns
Coca-Cola used generative AI to recreate its famous “Holidays Are Coming” advertising style in 2024 and returned with another AI-produced version in 2025.
The campaign showed both the opportunity and the risk of applying generative AI to emotionally important brand assets. AI allowed Coca-Cola to produce complex winter environments and visual sequences efficiently, but the 2024 campaign also attracted criticism over unnatural-looking people and the synthetic recreation of a nostalgic advertisement.
The 2025 version reduced the emphasis on realistic human characters and produced more refined visuals.
How AI was used:
Generative video tools were used to create scenes, characters and environments associated with Coca-Cola’s holiday advertising.
Why it matters:
The campaign demonstrated that technical quality is only one part of successful AI advertising. Audience expectations matter just as much.
What marketers can learn:
Fully synthetic creative carries more risk when a campaign depends on nostalgia, recognizable human emotion or an existing piece of brand heritage.
3. Popeyes’ AI-Generated Wrap Battle
When McDonald’s brought back its Snack Wrap, Popeyes responded with an AI-generated campaign built around a rap battle between the competing products.
The team used Google’s Veo 3 for video generation and Suno for AI-generated music, allowing the campaign to move from idea to finished creative in less than three days.
How AI was used:
Generative AI produced both the video and music, dramatically reducing the production time normally required for a reactive campaign.
Why it worked:
Speed was part of the creative strategy. Popeyes could respond while the competitor announcement was still culturally relevant.
What marketers can learn:
One of generative AI’s strongest advertising advantages is reactive production. A timely idea that traditionally takes weeks can sometimes be produced while the conversation is still happening.
4. Mango’s Fully AI Fashion Campaign
Mango used AI to create campaign models, backgrounds and lifestyle scenes for its Teen collection while keeping the actual clothing grounded in real product photography.
The campaign ran across 95 markets and demonstrated how generative AI can replace parts of a traditional fashion shoot without replacing the real product being advertised.
How AI was used:
Product photography acted as the foundation while generative AI created models and environments around the clothing.
Why it worked:
AI solved a genuine production problem: generating large amounts of lifestyle imagery without organizing multiple models, sets and locations.
What marketers can learn:
Generative AI is particularly useful when the product itself must remain accurate but everything around the product can be produced synthetically.
5. Under Armor’s Anthony Joshua Commercial
Under Armour used generative AI as part of its creative production process to work around practical limitations involving traditional filming and talent availability.
Rather than treating AI as the campaign idea itself, the technology became another production tool used to construct the final creative.
How AI was used: Existing creative material could be extended, adapted, or recreated through generative production techniques.
Why it mattered: AI reduced some of the constraints normally associated with organizing new shoots and recreating complex scenes.
What marketers can learn: Generative AI can provide the most value when it removes a specific production limitation rather than replacing the entire creative process.
6. Toys R Us’s Sora-Generated Brand Film
Toys R Us introduced a brand film created with OpenAI’s Sora to tell the story of its founder and mascot. While the campaign gained significant attention, audience reactions were mixed. The example highlights that fully AI-generated storytelling can still face challenges, especially when audiences have strong emotional connections with a brand.
Toys “R” Us experimented with generative AI video to tell a story connected to the history of the brand.
The campaign used AI-generated environments and characters to create scenes that would otherwise have required conventional sets, actors, visual effects, or animation.
How AI was used: Generative video helped create the visual environment and storytelling sequences.
Why it mattered: The campaign demonstrated that generative AI can support longer-form brand storytelling, not just short performance ads.
What marketers can learn: AI-generated storytelling works best when the technology supports a clear narrative. When the synthetic nature of the output becomes more noticeable than the story itself, it can distract from the campaign.
7. JPMorgan Chase’s AI-Generated Ad Copy
JPMorgan Chase used AI to improve advertising copy and experiment with messaging variations.
Unlike many well-known generative AI campaigns that focus on video or images, this example highlights one of the most practical applications of AI in advertising: producing and testing different ways of communicating the same value proposition.
How AI was used: AI supported the creation and testing of advertising copy variations.
Why it mattered: Copy testing can be scaled more easily than traditional brainstorming and manual variation creation.
What marketers can learn: Generative AI advertising does not need to involve synthetic video or imagery. Testing headlines, hooks, benefits, objections, and CTAs can be one of the lowest-risk ways to use AI in advertising.
8. H&M’s AI Digital Twins
H&M created AI-powered digital twins of real models to produce marketing images without organizing full photoshoots. This allowed the brand to test different outfits, poses, and lighting quickly while reducing production time and costs.
H&M has explored AI-generated and digitally recreated models as part of its approach to fashion content and creative production.
Digital representations can give brands more flexibility in how existing talent appears across markets, environments, campaigns, and creative formats without requiring a separate physical shoot for every asset.
How AI was used: AI-supported digital models and environments can be used to produce additional campaign creative from existing talent and product assets.
Why it matters: The approach creates new production flexibility but also raises important questions around consent, authenticity, ownership, and how synthetic talent should be presented.
What marketers can learn: AI-generated people require much stronger governance than background generation or creative variation. Brands need clear permission, quality control, and transparency processes before scaling this approach.
What the Best Generative AI Advertising Examples Have in Common
There are a few clear patterns once you look across all of them, and they hold whether the budget is seven figures or a fraction of that.
| Principle | What It Means | Example |
| AI solves a specific problem | The technology improves production, personalization or speed rather than being added for novelty | Under Armour |
| The creative idea comes first | A clear campaign concept gives the AI output meaning | Nutella |
| Human direction controls quality | Creatives still decide narrative, pacing, brand fit and final output | Mango, H&M |
| AI enables faster production | Generative workflows can make reactive or high-volume campaigns possible | Popeyes |
| Brand context matters | Audiences tolerate synthetic creative differently depending on the category and emotional context | Coca-Cola, Toys “R” Us |
| Testing matters more than generation | Producing more variations only helps when brands measure what performs | JPMorgan Chase |
The common pattern is that successful AI advertising does not begin with the question, “What can we generate?” It begins with a marketing or production problem and asks whether generative AI provides a better way to solve it.
Generative AI Advertising Ideas Smaller Marketing Teams Can Actually Use
Most of the use cases above translate directly to a mid-sized brand’s marketing team, once you strip away the scale:
- Product visual variation, like Nutella’s jars or Mango’s AI-generated lifestyle imagery, works at any scale. A DTC brand can generate dozens of packaging or product image variants to test which resonates before committing to a full print run.
- Script and copy generation with human editing, like the JPMorgan Chase example, is the lowest-lift starting point. AI drafts; a real writer or creative director edits.
- Video without a full production crew, like H&M’s digital twins, is where the biggest cost savings show up. It is also where the risk of the Coca-Cola and Toys R Us problem is highest: without a filmmaker’s eye on the final cut, AI-generated video can look uncanny instead of polished.
That last point is the real gap between the brands above and everyone else. Generating AI video is now accessible to almost any team. Making it look like it came from a studio, not a prompt, still requires people who know how to direct, edit, and judge a shot the way a filmmaker would.
Enterprise campaigns may use custom AI systems and large production budgets, but the underlying approaches are accessible to much smaller marketing teams.
1. Generate Creative Variations From One Winning Concept
Instead of producing ten completely different campaigns, start with one strong idea and use generative AI to test different:
- Hooks
- Opening scenes
- Calls to action
- Backgrounds
- Product angles
- Aspect ratios
The goal is not variation for its own sake. Each version should test a clear hypothesis.
2. Create Product Lifestyle Imagery Without a Full Photoshoot
Brands can keep real product photography while generating different environments, models or use cases around it.
This approach is particularly useful for ecommerce teams that need fresh campaign visuals more frequently than traditional photography budgets allow.
3. Produce Short-Form Video Without a Full Production Crew
AI video tools can help create product demonstrations, visual concepts, animated scenes and campaign cutdowns without organizing every scene through a physical shoot.
Human editing is still important for continuity, pacing and brand quality.
4. Use AI for Copy and Hook Testing
Generating headline variations remains one of the lowest-risk applications of generative AI advertising.
Marketers can test different:
- Benefits
- Pain points
- Objections
- Offers
- Calls to action
The winning messages can then inform future creative.
5. Localize Existing Campaigns
Generative workflows can help adapt existing creative across languages, markets, presenters and formats without rebuilding the campaign from scratch.
This makes AI especially valuable for brands running the same offer across multiple markets.
When Should Brands Be Careful With Generative AI Advertising?
Generative AI is not automatically the right production method for every advertisement.
Brands should be especially careful when a campaign depends heavily on:
- Real customer testimony
- Human authenticity or emotional storytelling
- Historically important brand imagery
- Accurate representations of people or products
- Sensitive social or cultural subjects
- Claims that require strict factual or regulatory review
Synthetic output should also go through the same brand, legal and quality-review processes as traditionally produced advertising.
Generative AI works best when it removes a production constraint without weakening the credibility of the message.
How Marmalaide Produces Generative AI Advertising Creative
Marmalaide combines generative AI production with human creative direction to create campaign-ready video and advertising assets.
The production workflow covers strategy, scripting, AI-assisted production, editing and campaign variations, allowing marketing teams to produce more creative without relying on a traditional shoot for every asset.
Teams that need ongoing advertising creative can explore Marmalaide’s AI ad creative production services, while campaigns focused specifically on video can use its AI video production services.
If your team wants to run its own version of these generative AI advertising examples, from AI avatar explainers to fully produced brand films, browse QuickFix for one-off projects or get in touch to talk through an ongoing content plan.
Frequently Asked Questions on Generative AI Advertising Examples
What are the best generative AI advertising examples?
The strongest examples pair AI-generated output, whether video, images, or copy, with real creative direction. Nutella’s personalized jar labels, H&M’s AI digital twins, and Mango’s fully AI-generated fashion campaign each succeeded because a human team shaped the final result rather than publishing raw AI output.
What are common AI use cases in advertising?
The most common use cases are image and video generation, ad copy generation and testing, personalized product recommendations, and reactive creative built around a real-time cultural moment. Brands are also using AI for creative testing, generating many ad variants quickly to see which perform best.
How do brands use generative AI in advertising without a large in-house team?
Most brands start with a single use case, often AI-generated ad copy or a single video product, then expand once they see results. Working with a production partner that pairs AI tools with real filmmakers and marketers, like Marmalaide, lets a smaller team access studio-quality output without hiring an internal AI creative team.
Why did Coca-Cola’s AI holiday campaign generate backlash?
Coca-Cola’s first AI-generated remake of its classic holiday ad, released in 2024, drew criticism for visuals many viewers described as unsettling, particularly around AI-generated human faces. The 2025 version addressed this by removing human depictions almost entirely, which performed better on quality benchmarks, though the underlying tension between AI and nostalgic brand storytelling remains unresolved.
Is fully AI-generated video creative safe to use for brand advertising?
It depends heavily on context. Fully AI-generated visuals have performed well in fashion and performance advertising, where audiences evaluate the ad on utility rather than authenticity. Nostalgic or emotionally loaded brand campaigns carry more risk, since audiences have clear expectations for how those moments should feel, and synthetic quality is more likely to be noticed and criticized there.
What made Popeyes able to produce an ad in under three days?
Popeyes used Google’s Veo 3 for video generation and Suno for AI music production, with an AI filmmaker scripting and producing the entire campaign. The team started with a slower image-to-video approach and switched entirely to Veo 3 once that proved too slow for the timeline, showing how tool selection matters as much as the underlying AI capability.
Do these generative AI advertising examples require an in-house AI team to replicate?
No. Most of the underlying capabilities- video generation, AI avatars, image generation, and script assistance- are available through production partners who combine the tools with a real creative team. A brand does not need to hire an internal AI department to apply the same thinking these examples demonstrate.
