In the latest edition of our series How AI Will Change, we spoke with ad and marketing tech veteran David Berkowitz about the myriad ways artificial intelligence is turning creative and personalization inside out. Berkowitz is the founder of the AI Marketers Guild and Serial Marketers communities, which were acquired by Marketecture Media earlier this year. He is currently serving as Marketecture’s Chief Community Officer.
As Berkowitz tells it, AI is collapsing traditional production barriers, turning every marketer into a potential creative powerhouse while fundamentally altering how brands compete for attention. The question is no longer whether AI will transform creativity. Rather, it’s about how quickly creative teams will adapt to a world where personalization happens at an infinite scale and the cost of producing compelling content approaches zero.
I don’t have the percentage, but this is just getting started, and it’s going to increase dramatically. The ways AI shows up in ads will range from very subtle to very obvious. Sometimes it’s just fine-tuning an element, the way Photoshop is used today. Other times it’s generating an ad entirely with AI. Over time, relatively few ads won’t be touched by AI in some form.
Creatives are still central to the process, and I hope that continues for the long haul. In the best-case scenario, AI gives great creatives more power—more tools, more options, and new ways to scale their work across formats and media. It also gives marketers who haven’t had the resources to scale, or even to advertise much, the ability to do far more. The role of the creative will change. In some ways, it’s about being more like an editor than a writer—recognizing great work, guiding the process, and making sure outputs match the vision. There’s still plenty of need for humans to be integral to the chain.
Global brands are managing creative across dozens of markets and hundreds of territories.
Localization and translation are big opportunities. Dubbing, for instance, has gotten remarkably good—not just seamless in the audio but with lip movements that look like it was re-shot in a new language. That’s improving by the day. But different brands have different tolerance levels. Some will insist on human review of everything, while others will be fine with “good enough.” For many, AI can already deliver localized or globalized content at the push of a button. The error rate will never be zero, but it will keep shrinking. It really comes down to risk and error tolerance.
Marketers now have far more options for creative validation. Some are controversial, like simulated audiences or synthetic data, and there’s pushback about how well those approaches mirror reality. Still, there are strong case studies showing they can reflect real-world results. Beyond that, AI massively scales what advertisers are already used to: testing lots of creatives with small sample posts for micro-audiences. What’s new is the exponential scale and the ability to detect patterns that humans wouldn’t program for—or even notice—in the first place.
It’s hard to generalize. We’re already seeing some powerful AI-generated short films, and even national TV ads, that trigger strong reactions. Tools are getting better at recognizing and expressing emotion, and research shows AI can even create memes that outperform human ones in terms of humor. So we shouldn’t dismiss AI’s ability to capture or trigger emotions. It may not invent something like Arrested Development or Curb Your Enthusiasm, but it can still generate content that moves people.
It’s a spectrum. Some marketers don’t care how creative the work is—as long as it generates clicks, awareness, or sales. For them, AI can deliver campaigns that are effective but not especially creative. At the same time, there’s a risk of soullessness. It’s hard to believe AI will ever rival Shakespeare, but people are fooled all the time by machine-written sonnets. That’s why we need to be thoughtful about when AI is just reducing friction versus when it’s replacing creativity—and why we also need to watch the emotional well-being of teams as AI gets more deeply integrated.
Personalization is one of the first areas where AI really shines. Off-the-shelf tools already let you take a message—say from a video avatar—and personalize it at scale. We’re seeing it in entertainment and celebrity campaigns, where instead of a generic greeting you get one tailored to you.
The incremental cost of personalization is so low that it becomes powerful. Most recipients won’t care whether it’s AI or not, especially as media literacy grows. If Messi or Khloé Kardashian sends a million “personalized” messages, people will realize it wasn’t done one by one—but it still feels special.
First-party data is like the new oil. That’s true for large language models and it’s true for brands. Controlling your own data lets you do more with it and stops you from being overly dependent on third parties. It’s not just about having more—it’s about being able to use it directly to create stronger, more personalized interactions.
It’s a mixed bag. Attention spans keep shrinking, which makes things harder. AI can flood the world with content, which means even more noise and more low-quality material that still manages to capture attention because it plays to the algorithm. That can erode the quality bar while still rewarding mediocrity. So while AI gives us new tools to compete for attention, it also creates more clutter that makes finding the signal even tougher.
I don’t know if prompt engineering is really a lasting skill—it’s just part of using tools that will keep evolving. The bigger shift is that smaller teams can be more nimble and take on more types of work. Larger teams will need to adapt too, but for everyone, it raises the bar for technological literacy. Every role now has to understand where AI contributes, where it doesn’t, and what the team wants to avoid automating. It’s about making those conscious choices.
We’ll still look at attention and effectiveness metrics, but there are new measures too. Share-of-voice inside AI systems — how a brand shows up in ChatGPT, Gemini, or AI overviews—is becoming a big one. It’s no longer just about marketing to humans but also to the bots that influence human choices. That’s a fundamental shift, because it requires a very different strategy.
Marketers always need to be careful about what first-party data they collect and how they use it. Campaigns can also be a way to gather more first-party data—through interactive experiences that feed back into CRM systems for future use. But consumer comfort levels are all over the place. Some people will upload tax returns or lab results to an AI tool, while others won’t even share their first name. So it’s a moving target. The key is to be respectful, play the long game, and avoid shortcuts — especially anything that risks leaking consumer data.
