Netflix, AI Agents and the Quiet Rubberizing of Advertising

Netflix has just given us one of the clearest signals yet of where advertising is heading.

At its latest upfront, the company said it is testing AI agents that can manage, optimise and purchase ads on Netflix. Not just generate copy. Not just suggest audiences. Actually run parts of the advertising process.

That matters.

For years, most AI marketing discussion has been stuck at the content layer. Faster blogs. Cheaper banners. More variants. Useful, but not transformative on its own. The bigger shift is operational. AI is moving from helping people make marketing assets to helping run the system those assets live inside.

Netflix is already talking about AI-driven media planning, creative adaptation for different formats, better matching between advertiser creative and Netflix content, personalised ad loads and dynamic frequency caps.

In plain English: campaigns that understand the viewer, the format, the context and the objective, then adjust accordingly.

That is a much bigger idea than “AI made me an advert.”

It points toward advertising that behaves more like a living system. Creative, media, targeting, optimisation and measurement all connected. Less manual handoff. Less spreadsheet theatre. More real-time adaptation.

I think this is where things get interesting.

The best marketing has always needed judgment. That does not disappear. Someone still has to decide the story, the audience, the commercial objective and the line you do not cross. But once that intent is clear, a lot of the machinery underneath can be handled by agents.

That is the part brands should be paying attention to.

AI agents will not just write headlines. They will brief variants, test audience segments, monitor performance, adjust spend, adapt formats, flag fatigue, tighten frequency, and eventually negotiate across platforms. The job of the marketer moves up a level. Less button-pushing. More orchestration.

Netflix is a useful signal because it sits at the intersection of entertainment, data, attention and advertising budgets. If a platform of that scale is testing agentic campaign operations, this is not a side experiment anymore. It is the direction of travel.

There is a risk, of course. If everyone uses the same systems in the same way, advertising gets even more bland. More optimised, but less interesting. More efficient, but less memorable.

That is why the human layer still matters. Taste matters. Brand judgment matters. The weird idea in the room still matters. “Rubberizing” definitely matters, even if nobody can quite define it yet.

The future is not fully automated marketing. It is sharper humans with agentic systems underneath them.

Netflix is showing the platform side of that future. The next phase will be brands asking how much of their own marketing operation can work this way too.

Not just faster content.

Campaigns that learn.

Media that adapts.

Creative that changes shape.

Agents that run the rubberizing while humans decide what should exist in the first place.

That is where this gets exciting.

Netflix source: Netflix Upfront 2026: Get Closer