When Ulta Beauty chief marketing officer Michelle Crossan-Matos discusses the importance of responsible use of generative artificial intelligence, she thinks of the moment children experiment with beauty care for the first time. And she wants to ensure the relationship they develop with beauty products is healthy.

“We’re specifically using gen AI to help speed up asset creation, but it is not to replace people,” says Crossan-Matos. “I don’t want [generative] AI to replace the authentic and natural look that comes from the real individual.”

While Ulta Beauty won’t base marketing on an AI-created face, there are plenty of creative ways the beauty retailer is using AI today. It is a tool that helps associates answer questions from customers, offers personalized products to users of the Ulta Beauty mobile app, and even for campaign ideation. Generative AI developed some visual concepts for a real-life photo shoot that Ulta Beauty conducted for a recent holiday campaign.

“AI has made its way across the company,” says Crossan-Matos.

She says the AI journey began in 2018, when Ulta Beauty acquired an AI-powered shopping assistant and an augmented-reality tech provider. That spurred future innovations like GLAMlab, which allows guests to virtually test makeup with their phone, and an AI-powered skin consultation tool called SkinAnalysis.

Ulta Beauty is ahead of the pack for the use of AI in marketing. A CMO survey conducted last year found that 94% of marketers only began to leverage the tech in the past three years. And 60% report beginning their AI journey less than one year ago.

Historically, AI was used by marketers for advertising targeting, segmentation for ads, email marketing, and audience targeting. But marketers are especially excited about the potential of more personalizing marketing, matching the exact content—text, images, and video—that will compel a customer to buy.

“For us, the reason why we keep growing our loyalty base is because of the content we serve,” says Crossan-Matos. “And you cannot serve content to 42 million loyalty members if you don’t have AI enabling you to tailor and personalize.”

Adrian Fung, CMO of eBay, says the e-commerce giant has over 130 million buyers and nearly 2 billion listings at any given time. The big challenge it faces: That’s a lot of content for eBay to sort through and ensure is shared with shoppers in a manner that’s compelling, delivered at the right time and the right frequency, and with the most optimal messaging possible.

“What I find exciting about AI is the ability to create that better customer experience,” says Fung. “There’s still a lot of exploration with generative AI and trying to figure out the best use cases.”

And eBay is experimenting in a few ways. Generative AI is helping it more finely tune banner ads and creating more personalized subject lines for email blasts. It is also helping users improve the image quality and descriptions of their listings on the eBay site. “We feel like we are still just scratching the surface,” says Fung.

Fung sees great efficiency gains with generative AI tools, but says art direction and creative ideas will remain driven by people.

“I think we at eBay, and a lot of marketers, are grappling with, what’s the interaction between those two?” asks Fung. “How do we blend that art and that science together, particularly as the science starts actually creating some of the art on its own?”

Marketers say AI will also help the industry become even more creative. Mastercard CMO Raja Rajamannar says marketers lost a lot of their creativity beginning in the mid-1990s, when marketing pivoted from being a creative function that was centered on psychology, design, and aesthetics, and instead became very driven on technology and data analytics.

“This is the opportunity I see with the advent of [generative] AI, in particular, that marketers have a fantastic opportunity to reclaim their time, because everyone else has access to the same set of technologies, the competing field becomes very level,” says Rajamannar.

With a more level playing field, creativity will become the ultimate differentiator. “This is the golden era we are about to enter for marketing,” says Rajamannar. “And if they leverage these tools appropriately, to bring the creativity forward, it’s going to be magical.”

The credit card company has used AI within the marketing function for more than six years. For Mastercard’s social media marketing, AI helps discover and even predict microtrends that can last for just a few days. AI is then quickly deployed to propel the communication messaging for those ads, the templates created for the optimal banner ad color and graphics, conducts testing to land on the final product, and makes the ad purchase. This entire end-to-end process is a combination of AI and automation and has helped greatly boost the effectiveness of Mastercard’s ads.

“I think marketing has kind of already lost creative creativity, sadly, in a lot of ways,” says Kipp Bodnar, CMO of HubSpot. “And if you believe that’s the case like I do, then you might say, ‘All right, is AI going to make that worse or is it going to make that better?’ I’m an optimist. I think it’s gonna make it better for a few reasons.”

HubSpot, which makes software for sales and marketing, a year ago assembled a small team of AI engineers and connected them with the marketing team to think of the best use cases possible in the wake of the generative AI breakthrough. They came up with 120 possibilities.

From there, HubSpot narrowed it down to roughly 20 and considered how valuable each use case would be, as well as how ready the AI tech was for those possible applications. “At any one time, we have three to six AI experiments rolling,” says Bodnar. And that’s just for marketing.

Marketers, he adds, do a lot of repetitive work. There’s a lot of content that must be created for a variety of channels. AI can help address the density of those asset needs. “I think that creativity is going to come back by giving marketers more time,” says Bodnar.

Duolingo CMO Manu Orssaud says AI is first and foremost a tool that helps increase productivity. With more than 50% of spending on performance marketing, getting more time to be creative is critical, because that’s where a marketer has more control.

“What we are really interested in around AI is the ability to help us optimize ourselves,” says Orssaud.

AI helps Duolingo discover social media trends that would resonate with the language app’s fans. When Netflix airs new seasons of the comedy series Emily in Paris, Duolingo’s marketing team works to capitalize on the spike in interest in French. K-pop fans tend to intersect nicely with Duolingo’s users, and the company has created content to build affinity with that audience as well.

“Because social [media] is so rich in terms of content and data, and sorting that data and making sense of it, I think this is where AI can play a massive role for marketing,” says Orssaud.

That said, AI should firmly remain a tool for marketers, not a solution for creativity.

“I’d rather have the designer focusing their time on thinking about concepts, thinking about ideas, thinking about how best to message something, and then AI can be the thing that helps us to scale and create variations,” says Orssaud. “AI doesn’t replace the creative process.”