Agentic AI could transform advertising production, by automating not just content generation, but the processes that enable its creation. In an op-ed, Xavier de Baillenx, CEO and co-founder of Ramdam, argues that the real challenge will not be to produce more, but to do so at scale without sacrificing creative diversity.
Just a few years ago, producing several dozen advertising videos required mobilizing teams, coordinating shoots, writing briefs, and managing numerous exchanges between brands, agencies, and creators. Artificial intelligence now promises to simplify a large part of this work.
Generative AI produces text, an image, or a video from an instruction. Agentic AI, on the other hand, coordinates multiple actions to achieve an objective, drawing on various tools and information sources. Applied to advertising, it can spot trends on social networks, propose creative concepts, prepare briefs, facilitate coordination with creators, and analyze campaign results.
This automation raises a question. If companies use the same technologies, the same data, and the same optimization mechanisms, aren’t they at risk of all producing the same advertisements? On social networks, where formats evolve quickly, replicating a recipe that worked can quickly reach its limits. An ad can respect an algorithm’s recommendations and yet go completely unnoticed.
The same brief, multiple interpretations
Take the example of a brand wanting to promote a fitness app. An AI system can identify several situations where the product adds value: getting back into sports, organizing workouts, progressing towards a goal, or regaining motivation. But entrusting the same concept to several creators can give rise to radically different content. One will choose humor, another will tell a personal story, a third will film a scene from their daily life.
For the author, artificial intelligence brings orchestration capability, creators bring their personality, sensibility, and knowledge of usage, while data then allows for measuring results. Rather than asking AI for fifty variations of the same ad, brands can thus coordinate fifty different creative interpretations.
A work organization set to evolve
Marketing teams and agencies could spend less time on operational coordination, and more on defining creative strategies and analyzing results. Their role will not disappear; it will evolve. They will still need to retain control over important decisions: which concepts to develop, what criteria to use to select creators, how far to automate the distribution and optimization of a campaign.
Brands will also need to ensure content quality, usage rights, advertising transparency, and message consistency. “Automating a process does not exempt one from defining its rules or controlling its results,” the executive reminds us.
The debate often pits advertising entirely produced by artificial intelligence against creation that remains essentially human. Xavier de Baillenx believes this opposition misses the real change. The more production becomes accessible, the more the ability to propose singular ideas becomes valuable. “Tomorrow, the most successful companies may not be those that have automated their creativity the most. They will be those that used automation to unleash more human creativity.”
