How AI is Changing Ad Creative and Campaign Optimization

A close friend of mine (someone who is very much against AI) was over at my house recently, scrolling on his phone. He started laughing and showed me a video of two cats fighting inside a backpack on the back of a motorcycle. Completely absurd. Completely convincing. He thought it was real.

I told him it was AI-generated, and he was genuinely surprised.

That’s where we are. The content is getting good enough to fool people, even the skeptics. But “convincing” content isn’t the same thing as meaningful content, and in advertising, that matters.

That cat video? That took intention and creative decision-making. Someone understood what would be funny, what would be absurd in just the right way, and built it. The AI was a tool in the hands of someone who knew what they wanted to make.

That’s not what’s running your ads right now.

What platforms are doing behind the scenes

Meta and Google have both moved aggressively toward AI-first ad systems, but with a very different goal. Meta’s Advantage+ has become the default starting point for campaigns, meaning automation is baked in before you touch a single setting. The platform’s creative enhancements (also on by default) can crop images, adjust brightness, animate still photos, add background music and reformat your creative for placements you never intended.

Google went a similar direction with AI Max, eliminating keyword targeting entirely and letting Gemini match your landing pages to user intent signals. The AI decides placements, audiences and optimization. You supply the assets and the conversion data.

TikTok’s Symphony suite has reportedly enabled brands to cut content production time by as much as 70%, allowing rapid testing across dozens of creative variants without expanding production teams.

The efficiency numbers are hard to dismiss. Google Ads Insights documented a 20% reduction in cost per click when using AI-driven campaign tools, and Meta’s benchmarks show return on ad spend improving by 30 to 35% with AI optimization.

Those are real numbers. But they come with a glaring distinction: these platforms have optimized for automation and speed, not for quality or authenticity. And that’s where all the AI slop comes in.

What we as marketers need to watch out for

The biggest risk isn’t that AI will produce bad creative once. It’s that it will produce mediocre creative consistently, and you won’t always catch it before it runs. A 2025 study by the Interactive Advertising Bureau found that more than 70% of marketers had already experienced an AI-related incident in their advertising (off-brand content, unexpected placements or copy that didn’t reflect their messaging).

AI isn’t removing us from the process. It’s just making it easier for us to lose control of it.

There are a few specific things worth keeping a close eye on:

• Auto-enabled settings that re-enable themselves. Some AI automation settings on Meta and Google can turn back on after a campaign is edited. You think you turned it off. You didn’t.
• Creative drift. When AI assembles headlines, images and descriptions dynamically, the combination it serves may not match your brand voice or pass compliance.
• Keyword cannibalization. Implementing AI Max can shift budget toward category keywords over branded keywords, disrupting campaigns with differentiated return-on-ad-spend targets.
• Weak inputs producing weak outputs. Thin landing pages, vague messaging and sloppy account structures become bigger liabilities when AI is making more decisions. The AI can only work with what you give it.

For industries we serve like healthcare, financial services non-profits, where compliance isn’t optional, the stakes are even higher. An AI that animates your still photo or rewrites your headline in a “more engaging” direction can create real regulatory issues. Fast and cheap doesn’t mean much when the ad shouldn’t have run in the first place.

AI has its limitations

Here’s what the benchmarks don’t capture: connection.

People don’t click ads because the algorithm found them at the right moment. They click because something in that ad feels like it was made for them. That requires empathy. It requires understanding not just who the audience is, but what they’re actually going through. AI can mirror patterns. It can’t feel the weight of a patient making a difficult healthcare decision, or understand why a manufacturing company’s legacy matters to the people who built it.

That’s the argument we keep coming back to. As we’ve written before, human-centered content still wins, especially in industries where trust is the product. AI can scale creative. It cannot manufacture authenticity.

The brands producing “AI slop” aren’t failing because they used AI. They’re failing because they handed the creative brief entirely to the machine and walked away.

The silver lining

When used correctly, by people who actually understand the platform mechanics, the audience and the brand, AI-assisted advertising produces results that would have been impossible to achieve manually two years ago. Testing several creative variants where teams previously tested a couple is now a realistic production goal. Audience signals such as customer lists, purchase history, and behavioral data are feeding smarter targeting than demographic buckets ever could.

The question isn’t whether to use AI. It’s whether the person managing your campaigns understands enough about the underlying systems to stay in control of them.

An agency that understands how Meta’s Advantage+ settings behave after an edit. That knows when AI Max is the right call and when it will cannibalize your branded spend. That can supply the creative inputs AI needs to perform, and catch the outputs that shouldn’t run.

The difference that matters

The cats-in-a-backpack AI? That took thought. Intention. Someone making creative decisions. The AI running your campaigns right now? It’s making different decisions entirely—optimizing for speed and scale, not for the story your brand needs to tell.

The technology that can fool skeptics and the technology that’s defaulting to mediocrity are not the same thing. And brands need to know the difference.

AI makes good strategy faster. It doesn’t replace the strategy. And it definitely doesn’t replace the people who know how to build it.

In short: the technology is moving fast, the platforms are defaulting toward automation, and the brands winning right now are the ones with knowledgeable hands on the controls.