Breef just published a piece on the AI marketing backlash that's better than 90% of the trade coverage on this topic — which, to be fair, is a low bar, but I'll take it. Coca-Cola's holiday ad felt like a hostage video. Toys "R" Us tried to Sora their way into a campaign and got laughed off the internet. McDonald's Netherlands ran a "creepy" AI holiday spot and pulled it before the ink dried. Meanwhile Spotify, Netflix, and Starbucks quietly run AI under the hood and nobody complains, because nobody notices — the work just works.

Breef's read: brands that make AI the pitch get punished. Brands that make AI invisible infrastructure get rewarded. Correct. Directionally solid. I'll co-sign that part.

But it stops right at the edge of the actual point, and I'm not going to let a good diagnosis walk out the door with an incomplete prescription.

It Was Never "AI vs. No AI." Stop Framing It That Way.

Here's the line Breef didn't draw: the failure isn't using AI. The failure is asking AI to have the idea for you.

There's a canyon-sized difference between using AI to create the idea and using AI to create with an idea you already had. One is a tool. The other is you outsourcing the one job that was never negotiable — having something to say.

When a brand types "give me a holiday campaign concept" into a model, they're not being efficient. They're admitting nobody in the room had a thought. What comes back is polished, competent, and emotionally dead — because the machine pattern-matched the shape of a story without ever having one. Audiences can smell that from a mile away, even when they can't explain why. That's not an AI-detection skill. That's a fake-detection skill, and humans have had it since before advertising existed.

Now flip it. A human has the idea — the concept, the tension, the punchline — and then brings AI in to execute, scale, personalize, produce. That's Spotify recommending your next song. That's Netflix picking your thumbnail. The AI never had the idea. A person decided what mattered, and the machine just moved faster than a room full of humans ever could. Nobody's hiding that AI's involved — it's just not the point, because it was never supposed to be.

The Part Breef Left on the Table: This Was Always a Storytelling Problem

This is where the article stops short, and it's the part that actually matters.

This isn't a tooling gap. It's a talent gap that AI ripped the curtain off of.

Marketers who can actually tell a story — who understand stakes, tension, payoff — don't reach for a model to invent their concept. They already have one. For them AI is a production accelerant: faster cuts, more variants, cheaper execution, zero compromise on the idea because the idea was never in question.

Marketers who can't tell a story are the ones handing the concept to a machine. And it shows up instantly, in every one of Breef's own examples, because a bad idea has always produced a bad ad — with or without AI in the pipeline. Weak concepts existed in 2005. AI didn't invent hollow marketing. It just removed the friction and the timeline that used to keep the laziest ideas from ever reaching production. Now they go live at scale, in days, and get dragged on the internet by lunch.

So no — the divide Breef describes isn't really about AI sophistication. It's about whether the humans running the brand have a story worth telling. AI didn't create that skill gap. It just stopped hiding it.

Bottom Line

Get the idea right first. With a human. Before AI touches a single frame. If your brand can't say what it's actually trying to say without a chatbot's help, no amount of AI polish is walking that campaign back from the internet's group chat. And if you can say it — if the story's already there — AI stops being a crutch and becomes exactly what it should be: leverage.

The brands getting torched right now aren't the ones using the most AI. They're the ones who never had anything to say and thought a machine could say it for them.