A client called me last Tuesday, mildly panicked. Their display ads had appeared next to AI-generated content that was, let's say, thematically inappropriate. Not offensive in the traditional sense. Just deeply, surreally wrong in a way that only AI-generated content can be. The kind of wrong where you stare at it for thirty seconds trying to figure out if you're having a stroke.
This is the brand safety challenge of 2025, and it's fundamentally different from the brand safety challenges of 2018 or 2021. It's not about your ad appearing next to hate speech or misinformation (though that's still a problem). It's about your ad appearing next to content that is technically "safe" but experientially bizarre. Content that was never meant to be seen by humans. Content that exists solely to capture programmatic ad dollars.
The scale of the problem is staggering. By some estimates, AI-generated content now accounts for a significant portion of new content published online daily. Much of it is what I'd call "gray content." Not harmful. Not valuable. Not created for any audience. Just... content-shaped material designed to fill space and monetize attention.
Traditional brand safety tools were built for a binary world: safe or unsafe. But we now live in a world where your ad can appear in a context that is technically safe by every keyword filter and sentiment analysis tool available, yet still makes your brand look absurd by association. The adjacency problem has evolved from "is this content offensive?" to "is this content even real?"
Here's my strategic take: the brands that will navigate this best are the ones shifting from programmatic reach to contextual precision. Yes, it costs more per impression. Yes, it limits scale. But every impression against AI-generated gray content is a wasted impression anyway, so the real cost comparison isn't reach vs. precision. It's quality attention vs. the illusion of attention.
But every impression against AI-generated gray content is a wasted impression anyway, so the real cost comparison isn't reach vs.
The second shift is from brand safety to brand suitability. It's not enough to avoid the bad. You need to actively seek the good. Positive adjacency, placing your brand next to content your audience genuinely values, is the competitive advantage in a landscape flooded with noise.
We're at the beginning of this problem, not the end. The content flood is only accelerating. The brands that recognized programmatic's limitations early will have a meaningful head start over those still optimizing for CPMs against audiences that might not be human.