Digital marketing in 2026 no longer resembles that of 2024. Search behaviors are fragmenting between traditional engines, conversational assistants, and social networks used as search engines. At the same time, the European regulatory framework imposes new constraints on the use of artificial intelligence in brand communication. These two simultaneous movements are reshaping the priorities for any company looking to maintain or improve its online visibility.
AI Transparency Obligations and Digital Marketing: What the European Regulation Changes
Since August 2, 2026, Article 50 of the European regulation on AI requires brands to inform individuals when they interact with certain artificial intelligence systems. The same provision mandates the disclosure of deepfakes that are publicly disseminated. For generative tools already on the market before this date, the obligation for machine-readable labeling has a deadline extension until December 2, 2026.
In practice, an advertising campaign using an image, video, or audio generated by AI that could be mistaken for an authentic representation falls under this disclosure obligation. The example cited by regulatory sources: a video generated of an executive broadcast by a marketing team without mentioning its synthetic nature.
This framework does not only concern large companies. Any brand that uses a chatbot on its site, that disseminates visuals produced by an image generator, or that publishes synthetic audio content must verify its compliance. The resources available around marketing on the Bordel de Nerd site allow tracking these regulatory developments and their concrete implications for digital strategies.
Synthetic advertising visuals require specific control before dissemination. An AI-generated visual presented as authentic exposes the brand to a real legal risk, not just reputational. International campaigns must also anticipate regulatory fragmentation between jurisdictions.
Conversational Search and GEO: Visibility Beyond Google
AI assistants (ChatGPT, Perplexity, Gemini) are changing the way users access information. An increasing portion of searches no longer goes through a traditional results page but through a conversational response that may or may not cite sources.
The term GEO (Generative Engine Optimization) refers to all practices aimed at having a brand or content cited by these assistants. Unlike traditional SEO, being cited by an AI becomes as strategic as ranking first on Google. The criteria that favor this citation remain partially opaque, but several elements recur in the available analyses:
- The structural clarity of the content (factual data, direct answers to questions, proper semantic markup) increases the chances of extraction by language models.
- The domain’s reputation and the frequency of citation by other sources play a role in the selection of generated responses.
- Content that provides original data (study, field feedback, numerical comparison) is more frequently cited than generic summaries.
Data is still lacking to precisely quantify the share of traffic diverted from traditional engines to these conversational interfaces. Traditional audience measurements do not always capture these interactions, complicating the assessment of the return on investment of a GEO strategy.
Authentic Content vs. Synthetic Fatigue: A Credibility Arbitration
The proliferation of AI-generated content has produced a measurable effect on audiences: fatigue towards texts, visuals, and videos perceived as artificial.
Imperfection and human signature are becoming markers of credibility again. Brand content that carries the voice of an identifiable person, with their own formulations and positions, generates more engagement than a text polished by a generative tool without editorial revision.
AI should not be excluded from content production, however. The challenge lies in editorial supervision: AI as a production tool, not as the final author. An article written with the help of a language model and then revised by a writer who injects field data and an original viewpoint poses no problem.
The same article published as is, without proofreading or enhancement, risks both algorithmic penalties and loss of reader trust.
Micro-Communities and Dark Social: Where High-Value Conversations Migrate
The most influential exchanges on purchasing decisions no longer occur on the public feeds of social networks. They are migrating to private spaces: WhatsApp groups, Discord channels, Slack communities, direct messaging on Instagram or LinkedIn. This phenomenon, often referred to as dark social, largely escapes traditional attribution tools.
Traffic from dark social appears as “direct” in analytics, which skews the evaluation of acquisition channels. A recommendation shared via private message leaves no traceable mark, yet it can trigger a visit and then a conversion.
For brands, investing in thematic micro-communities around their sector yields more sustainable results than the race for impressions on public feeds. The logic changes: nourish a small group with high-value content, which will then be relayed by its members in their own private circles.
- Identify the spaces where the target already exchanges (specialized forums, private groups, professional channels) rather than creating an artificial brand space.
- Produce shareable content privately: exclusive data, operational summaries, downloadable practical tools.
- Measure impact through indirect indicators: spontaneous mentions, organic growth of direct traffic, conversion rates of “no source” visitors.
Digital visibility in 2026 is no longer just about ranking in a search engine or the number of followers on a social network. Regulatory compliance on AI, the ability to be cited by conversational assistants, and presence in private spaces form three complementary axes. Companies that arbitrate between these levers based on their actual audience will consolidate their visibility over time.
