What happens when AI designs ads?

The real question for publishers is no longer whether AI will play a role in advertising production. It’s how to use it intelligently.”

Kurt Jackson, SCS's Owner and Managing Member

Can AI actually design advertisements that publishers can use today?

Recently, we decided to find out.

We tested a range of AI tools with a simple challenge: create real advertisements that could run on a publisher’s website or appear in print. The lineup included Adobe Firefly, ChatGPT, Copilot, Canva, Gemini, Reeve and several other platforms.

The results were mixed, and revealing. Some ads looked polished and nearly production-ready. Others missed the mark entirely. What became clear very quickly was that success with AI wasn’t about the tool itself. It was about how the tool was used.

Prompting isn’t a step, it’s the process

One of the most important lessons from this exercise was that prompting isn’t simply part of the workflow. In many ways, it is the workflow. One of the most effective approaches we discovered actually reverses the traditional process.

Instead of starting by asking the AI to create an ad, we first asked it to generate a detailed prompt based on the ad specifications — size, audience, tone, message and format. That prompt was then used to generate the creative. This two-step method consistently improved the results. In many cases, the ads produced were usable or very close to it.

For publishers experimenting with AI-generated creative, this simple shift can make a significant difference.

See the ads used in this study

There is no single “best” AI tool

Another takeaway from our testing is that there isn’t one platform that consistently outperforms all others. Some tools are stronger at layout and visual composition. Others excel at generating marketing copy or conceptual ideas. A few do a good job of blending both. But none of them dominate every category.

For publishers producing creative for both print and digital platforms, this reality creates both a challenge and an opportunity. The challenge is obvious: the landscape is fragmented. The opportunity is that each tool can contribute something valuable to the creative process.

A flexible AI workflow

Rather than committing to a single AI platform, a more effective approach may be to build a flexible workflow that leverages several tools.

That means:

• Using different models for different tasks
• Comparing outputs side-by-side
• Selecting the best results from each system

This kind of hybrid approach provides more control over the final product and often leads to better outcomes.

From experiment to production

AI-generated advertising is no longer theoretical. Many publishers are already experimenting with ways to integrate it into real production workflows. The promise is compelling: faster turnaround, scalable creative production and reduced costs.

But as our experiment demonstrated, the technology still requires thoughtful oversight and a willingness to test different approaches. The real question for publishers is no longer whether AI will play a role in advertising production. It’s how to use it intelligently. And if our experience is any indication, the future of ad creation won’t be powered by a single AI tool — but by a smart combination of several.

About the author

Kurt Jackson is a 40-year newspaper industry veteran and the owner and managing member of Software Consulting Services (SCS). SCS provides publishing applications used by more than 500 sites in 10 countries producing over 16,000 publications in five languages.