With generative AI, media visibility now depends as much on the responses produced by conversational assistants as on search engines.
For more than twenty years, the media have learned to master search engine codes. Natural referencing has become a discipline in its own right, influencing both editorial strategies and business models. But a new transformation is underway. With the rise of generative AI like ChatGPT, Gemini, Claude or Perplexity, Internet users are no longer content with searching for links: they are waiting for answers to specific questions.
This development could seem like just a new stage in the history of the web. However, it constitutes a much more profound change. Because for the first time since the advent of search engines, the media are seeing the emergence of a new intermediary between their content and their audiences.
From research logic to conversation logic
For years, visibility was based on relatively simple mechanics: the user formulated a query, a search engine displayed a list of results to browse themselves and the media competed for the first positions.
Today, AI response engines are changing this logic. When a user queries ChatGPT, Gemini, or Perplexity, they first get a summary. Links to sources still exist, but they are no longer necessarily the main entry point to information. Above all, this summary becomes the starting point of a conversation: the user can delve deeper into a subject, ask for clarification, reformulate their query or explore new angles without leaving the interface. The information journey is no longer based on navigation between several contents, but on continuous interaction with a response that is enriched over the course of the exchanges.
This development is not trivial and its first effects are already visible. According to data relayed by the Reuters Institute and Chartbeattraffic from Google Search to news sites fell by around 33% in 2025. At the same time, media executives surveyed by the Reuters Institute anticipate an average drop of 43% in traffic from search engines over the next three years. As people get answers directly from AI-enriched search engines or conversational assistants, the historical link between viewability and click-through is weakening. For publishers, the challenge is no longer just to appear in search results, but to continue to exist in the environments where answers are now produced.
In other words, the battle for visibility is no longer fought only on the results pages. It also plays out in the AI response engines themselves.
Editorial credibility becomes a strategic asset
Faced with this new reality, not all content is equal. Generative AI favors sources deemed reliable to produce their answers. Original surveys, exclusive analyses, proprietary data or sectoral expertise thus become particularly valuable assets.
This development favors the emergence of new partnerships between media and actors of theartificial intelligence. In France, Le Monde entered into agreements with OpenAI then Perplexity to allow the use of its editorial corpus with attribution and remuneration. Internationally, OpenAI has also established partnerships with several major press groups, including The Associated Press, the Financial Times and The Atlantic. The same logic is required: in a world where response engines become gateways to information, the media’s ability to be cited, attributed and remunerated becomes strategic.
The challenge is all the more important as uses evolve rapidly. According to the Reuters Institute, 15% of under-25s use an AI assistant every week to get their news. The visibility of a media outlet could thus gradually be measured by its ability to be cited in the responses generated by AI, in the same way that editorial offices today monitor their natural referencing or their organic traffic. This development also opens up new measurement challenges. Tomorrow, media outlets and brands will not only need to track their visibility in search engines and traditional media, but also understand how their content is picked up, cited and recommended by conversational assistants.
Visibility, a new challenge for traditional media
After search engines and then social networks, generative AI introduces an even more profound disruption: it redefines the way in which content is discovered, interpreted and consumed.
After SEO and mastering social networks, publishers must learn to master their presence in AI response engines. And in this new attention economy, being the source of a response will be as important as being the destination of a click.
