Companies like hospitality chain Wyndham Hotels & Resorts, consumer healthcare firm Opella and betting brand BetMGM have little in common. What they do share is a broad portfolio of individual products and a need to spin up large quantities of digital creative assets to market them.
All three brands are building in-house creative teams trained to leverage AI to mass-produce digital creative assets, paired with media performance data that informs the next campaign or iteration. It’s a sign that scaled AI creative production is becoming ordinary among marketing’s upper middle tier advertisers.
While the first advertisers to press ahead and scale up AI production efforts were companies with industry-leading media and marketing budgets like Unilever, Coca-Cola or L’Oréal, advertisers outside that pantheon have begun to build out similar capabilities.
Several large advertisers use Brandtech Group’s Pencil platform as the backbone of the AI creative systems. While Brandtech has several Pencil clients bringing in over $10 billion in revenue, Pencil CEO Will Hanschell said its fastest-growing client segment was clients with less than $1 billion in annual income. “They culturally move faster,” Hanschell explained. “They’ve experimented faster… they’re looking for something bespoke, but less rigid than a traditional enterprise solution.”
Wyndham’s overall marketing and loyalty budget came to $565 million last year, according to its full-year results — hardly small potatoes, but dwarfed by the multi-billion dollar budgets wielded by AI first movers like Unilever or L’Oréal, which spent $1.98 billion on AI investments alone last year. The hotel chain has been working with AI firm Adora for a year to generate assets for use in its paid and owned Instagram activity, including video and still images.
“It allows us to showcase the breadth of our portfolio around the world, and ensure we are showcasing the right assets to an audience who’s looking for all-inclusive versus the urban city destination,” said Mike Shiwdin, gvp of loyalty, guest engagement and strategic partnerships at Wyndham.
Wyndham’s five person in-house marketing team uses the software to rapidly edit and create multiple versions of assets based on real images of its hotel estate. They also use the platform to monitor and receive feedback regarding how each asset performs on Instagram, combining data from media mix modeling and Meta’s platform API.
“Over time you’ll see performance go up becuase we’re figuring out what are the best performing creatives, and honing in on those,” said Adora CEO Marco Matos.
Per Shiwdin, it’s led the team to generate 15 times the amount of assets they were putting out before, and cut the hundreds of hours spent concepting, and the weeks required for production and approval required to launch campaigns, by 75%.
Furthermore, Shiwdin told Digiday that the company had seen customer engagement rise significantly since it began using the plethora of AI imagery, while it was increasing its media spend on Instagram; he didn’t provide financial specifics.
“We’ve seen a six times improvement in site visit rate over our benchmark,” he said.
Other companies have pursued their AI strategy entirely in-house. Consumer healthcare company Opella, which was spun out of Sanofi in 2025, has an 85-strong team using generative AI tools to write strategy briefs, and generate and animate marketing assets, including brand mascots.
Current iterations of allergy treatment Xyzal’s mascot — an owl called Nigel — are AI, according to Julien Leterrier, global head of martech at Opella.
Leterrier told Digiday the company was now producing “20 times” the amount of creative assets it produced prior to using AI tools, mostly variants or core assets produced for different languages and markets. “We’ve produced more than 20,000 pieces of content. We could not reach that scale before,” he said.
BetMGM, meanwhile, has pursued a more limited rollout. The company spends around $160 million on media annually, and has an established in-house creative department of 120 people. While it’s using AI tools including Arcana AI, a generative AI suite combining access to several models, to create imagery and video spots, some in-house production is reserved for human staff, given its strictly regulated category.
“That’s not to say humans aren’t capable of making mistakes, but we haven’t found that trust in the machine yet to actually do it all,” said Victor Gonzalez, senior director of creative at BetMGM.
All three companies still work with creative and media agencies; Wyndham’s AOR, for example, is Havas. But they’re assigned briefs that involve larger-scale, more traditional creative (like BetMGM’s recent Jon Hamm-starring above-the-line campaign) and which benefit from an outside perspective.
Brands in the industry’s upper-middle echelon have been able to adopt these capabilities, in part, due to the growth of platforms like Pencil or Wyndham’s partner Adora, that allow brands to outsource the technical lift of an AI creative setup. Meanwhile, the upfront investment required to build up such a system has fallen as compute costs have shrunk even as AI spending has risen; the price American businesses pay per million tokens fell 41% from its peak in March of this year.
There’s also the fact that generative creative tools have become more reliable for marketers concerned about visual consistency, noted Octavio Maron, chief creative partner, creative innovation at Dentsu.
“It’s become simpler for a broader range of clients to see the possibilities of how [AI] can be applied to their brands,” he said.
Then there’s the arms race factor. CMOs are obliged to keep up and make sure that category leaders aren’t the only ones able to deploy a greater number and variety of creative in their messaging. Four out of five (81% ) CMOs expect they will need to produce “significantly more content” in the near future, per a recent Dentsu survey.
As AI usage speeds up the production process, more firms are choosing to test creative before it goes out the door. The number of early-stage creative tests run by Kantar, for example, rose 17% between December 2025 and July of this year. The increase is a direct consequence of more brands embracing AI creative production, according to Kristi Rodgers, president for global solutions & product at Kantar.
“New qualitative and quantitative approaches are helping advertisers to identify the strongest creative routes before AI workflows take over, production begins and budget is committed,” she said in an email.
Wyndham, Opella and BetMGM aren’t embracing AI without reservation. BetMGM had held back from using the tools to generate assets carrying information about betting odds, for example, according to Gonzalez.
“You can’t hold AI accountable,” noted BetMGM’s Gonzalez. While BetMGM “Isn’t 100% there yet,” Gonzalez noted its stance may change as tools are proven reliable.
Leterrier, meanwhile, said Opella had put firm rules in place around depicting medical professionals, while appending AI disclosures to campaign materials running in markets that require them, such as the European Union. “No fake doctors,” he said.
Wyndham, Shiwdin said, was still debating whether or not it was appropriate to generate images of people within assets showing its hotels and resorts.
As trust in the tools among marketers improve, expect to see advertiser takeup of assembly line-style AI applications bloom.
More in Marketing
‘It scared the hell out of me’: What brands have learned from creator partnerships
At Shoptalk Fall, brands like Favorite Daughter, SharkNinja and Bob’s Discount Furniture talked about what they’ve learned from working with creators.
Future of Marketing Briefing: Agency fees in creator deals are the next transparency headache for marketers
As creator marketing budgets grow, so do questions about what agencies are taking.
Digiday’s guide to what’s in and out for Advertising Week NY 2026
Here are the trends as we see them for AWNY 2026.