The autumn months of 2026 are shaking up the marketing, media, and communication sector. Three underlying trends are reshaping the rules of the game: advertising is settling into conversational assistants, media quality measurement is adapting to these new environments, and audiovisual pluralism is subject to strengthened regulation in France. Here’s what to take away from these recent marketing and communication updates.
Advertising in ChatGPT: independent measurement changes the game
OpenAI is testing an advertising format directly integrated into ChatGPT. This is not a classic banner slipped into a chat: the advertisement appears in the conversational flow, linked to the context of the question asked by the user.
Have you ever noticed that a recommendation in a conversation seems more credible than an ad on the side of a web page? This is precisely the lever that conversational advertising in ChatGPT aims for. The format brings the product suggestion closer to the moment when the user is seeking an answer, without visual disruption.
The decisive point is not the format itself, but what accompanies it. OpenAI has partnered with DoubleVerify for campaign attribution and Integral Ad Science (IAS) for measuring brand safety and brand suitability. In practical terms, this means that independent third parties evaluate whether the advertisement appears in an appropriate context for the brand, and whether it generates a measurable effect on user behavior.
To follow all the news on CB News, this type of partnership marks a turning point: advertisers now have verification tools comparable to those of web display, applied to generative environments.
Brand safety in conversational environments
Brand safety in a chatbot does not work like on a website. On a classic page, categories of content (violence, misinformation) are excluded. In a conversation, the emotional context of the user comes into play. A person describing a moment of personal vulnerability should not receive a poorly calibrated advertisement.
The pilots announced around ChatGPT evaluate the conversational context in real-time. Some sensitive environments are automatically excluded. IAS tests suitability (brand-context appropriateness) with tools designed for this format, not simply transposed from the web.
Integrated commerce in AI assistants at Meta
Meta is pushing the integration of artificial intelligence across the entire advertising chain. The goal goes beyond simple automation of targeting or visual creation.
What is emerging is a complete circuit within a single interface: the AI assistant recommends a product, targeting refines in real-time, the advertising visual is generated automatically, and payment can be made without leaving the conversation. Recommendation, targeting, creation, and payment converge in a single flow.
For an advertiser, this reduces the number of steps between discovering a product and making a purchase. For the user, the boundary between personalized advice and advertising becomes blurred. This convergence raises a strategic question for brands: how to maintain trust when commercial recommendations borrow the codes of natural conversation?
GEO and visibility in conversational engines
You type a question into a classic search engine and get a list of links. With conversational engines (Google AI Overviews, Perplexity, ChatGPT Search), you get a synthetic answer that sometimes cites sources. Being visible in these AI-generated responses is the principle of GEO (Generative Engine Optimization).
GEO is becoming a structured commercial offering. Some media agencies are associating their editorial brands with generative advertising devices, meaning that the content from their media serves as a cited source in AI responses, with an integrated monetization aspect.
What GEO changes for a content strategy
Traditional SEO optimizes a page to appear in a list of results. GEO optimizes content to be selected as a reliable source by a language model. The criteria differ:
- The structuring of data (markup, citations, verifiable facts) weighs more than keyword density
- The reputation of the source domain influences the likelihood of being cited in a generated response
- Long and comprehensive content is favored over short conversion-oriented pages
GEO does not replace SEO; it adds a layer of visibility in interfaces where the user no longer clicks on a link but reads the answer directly.
Audiovisual pluralism and speaking time: what changes before 2027
Since October 2026, a new rule governs the speaking time of political figures in French audiovisual media. Arcom (the Regulatory Authority for Audiovisual and Digital Communication) now applies control over manifest imbalances in political representation on air.
Radio France, TF1, and CMA Media have requested more flexibility from Arcom in applying this rule. The tension is real: newsrooms must respect a quantified balance while retaining editorial freedom over the choice of guests and topics covered.
This regulation is part of the preparation for the 2027 presidential election. French audiovisual media anticipate a period of increased constraints on the distribution of airtime, with possible adjustments over the months.
- Arcom controls manifest imbalances, not strictly equal speaking time for each party
- TNT channels will have to integrate these constraints into their editorial grid from the pre-campaigns
- The call for TNT candidacies 2027 launched by Arcom adds further uncertainty to the French media landscape
The coexistence of these marketing and media updates outlines a rapidly changing landscape. Advertisers must navigate advertising formats that integrate into AI conversations, measurement metrics still in development, and a tightening audiovisual regulatory framework. Keeping track of these developments over the weeks remains the best way to adjust your communication strategy before it becomes outdated.
