Rethinking personas in the era of generative artificial intelligence

The notion of the persona is showing its limits: generative AI upends that logic. Personalization no longer rests on a fixed profile, but on a dynamic reading of language, context, and intent.

The persona facing the arrival of AI

Ask the same question on ChatGPT and two users will often get different answers. For good reason: AI does not answer a question. It answers a context.

It takes into account the history of the exchange, the tone used, the phrasing chosen. Each interaction becomes unique.

How AI redefines the idea of a target

Models can condense that history into "soft prompts" in order to build an implicit conversational frame. They rely on weak signals — turns of phrase, lexicon, style — to model implicit preferences.

Faced with this new reality, how do you structure a body of content able to embrace that variability? You have to think in terms of styles, postures, and conversational scenarios.

Toward a site as a conversational database

The persona becomes a cloud of possibilities, shaped by the moment and by the way one interacts with the machine.

This shift invites brands to turn their site into a conversational database: an ecosystem of highly segmented content, designed to be called, recombined, and reformulated by an AI.

Recognizing AI as a "meta persona"

AI becomes a filter, an actor. A discursive interface. This meta persona is not a target to reach, but a channel to understand.

For companies, that implies a double challenge: - Optimize their presence in the corpora that AI draws on - Anticipate the way their messages will be translated

AI becomes a co-author of the experience.

The persona becomes a moving point of balance, between the user, language, and the machine. Brands have to articulate three levels: the profile, the discourse, and the AI channel.