The risk of AI in branding isn’t that people are using it. It’s that too many are using it without thinking.

Used well, AI can open directions faster, pressure-test thinking, and accelerate the exploratory work that used to take days. But a tool is only as good as the judgment behind it. And right now, too much of that judgment has gone quiet.

The same prompts generate the same structures, produce the same outputs, at speed, and because it looks finished, it often escapes the much-needed scrutiny. Voice is flattening, style is disappearing, and yet output is coming faster than ever, but with less interrogation than ever.

This is where human judgement absolutely matters: knowing what the brand should refuse, when a sentence sounds like everyone else’s, what the audience is tired of hearing, and when the obvious answer is actually wrong.

We use AI in our process, most agencies do. The call isn’t to use it less (unless you want to get into the environmental repercussions of it all, but that’s another post), but to use it with more strategic discipline, sharper filters, stronger taste, and the willingness to challenge what it gives back rather than ship the first thing it spits out. Because if every brief runs through the same tools, trained on the same data, steered by the same instinct to reach for what’s fast and plausible, it’s the market that loses.