Creating an ad can be tricky and costly, and more brands see generative AI as a means to make that process less painful.

Reuters reported large advertising agencies like WPP and multinational corporations like Unilever look to generative AI to cut marketing costs while making more ads. 

WPP clients Nestlé and Mondelez, makers of Oreo and Cadbury, used OpenAI’s DALL-E 2 to make ads. One ad for Cadbury ran in India with an AI-generated video of the Bollywood actor Shah Rukh Khan inviting pedestrians to shop at stores. WPP’s CEO told Reuters savings from generative AI can be “10 to 20 times.”

Unilever created its own generative AI advertising tools and used them to write spiels for one of its shampoo products.

But as more brands look to use generative AI to make more ads, the question of whether they need to let people know it’s AI-generated rears its ugly head. Big AI companies signed an agreement with the White House to develop watermarking technology to label something made with AI tools. In theory, AI-generated ads could fall under that rule.

Ken Sickles, chief product officer for digital watermarking company Digimarc, tells The Verge that there should be standard policies on tagging AI-generated content.

“Marketing content is technically original AI content, and context is important when looking at something,” Sickles says. “Tagging that content is great because it brings context, but there has to be an industry standard policy around it.”

Marketing and advertising were part of the early wave of generative AI adopters, using language models to write copy — the words in an ad convincing you to buy the same coffee cup your favorite celebrity is holding. Generative AI copywriting tool Jasper became one of the darlings of the tech world in 2022, when it was reported the company raised $125 million in Series A funding.

And online brands also began experimenting with generative AI. For example, direct-to-consumer lingerie company Adore Me played around with the technology to figure out how to market bras and other intimate apparel without dealing with censorship. However, it still needs to put that into action. 

Companies using generative AI for advertising told Reuters they still have concerns over whether their AI ads can receive copyright protections and security risks when using AI.