A marketing team is launching a personalized email campaign using a generative AI tool. Open rates are soaring, but the legal department raises a warning: the chatbot integrated into the site does not yet comply with the European AI regulation requirements. This type of tension between performance and compliance aptly summarizes the daily reality of businesses in 2026.
AI Compliance and Marketing: What the European Regulation Changes in Practice
There is much talk about artificial intelligence as a lever for personalization. There is less discussion about the regulatory constraints that accompany its deployment in customer journeys. The European AI regulation (AI Act) notably requires informing users when they interact with certain AI systems and making identifiable certain generated or manipulated content.
For a marketing team, this translates into specific actions: clearly labeling a chatbot, indicating that an advertising visual was generated by AI, documenting the systems used. Transparency about AI becomes a matter of compliance, not just a trust argument.
The regulation also distinguishes between obligations that are already applicable and deadlines that have been postponed. To establish a realistic compliance timeline, it is beneficial to precisely check which provisions currently apply to marketing tools and which have a grace period.
This sorting work, often delegated to the legal department, should also involve marketing managers and data teams, as they are the ones configuring the tools on a daily basis. A comprehensive overview of the characteristics of marketing on Europe Entreprises helps to situate these issues in a broader context.
Retail Media and Incrementality Measurement: The Real Barrier for Advertisers
Retail media networks are experiencing strong growth. Advertisers see it as a valuable channel because it targets buyers at the moment of purchase. In practice, the deployment of budgets faces a measurement problem.
Due to the lack of common standards between retail media platforms, it remains difficult to compare the performance of one brand to another. The challenge lies especially in distinguishing sales simply attributed to an advertisement from those actually generated by it, known as incrementality. IAB Europe is working on establishing standards to harmonize these measurements, but the project is ongoing.
In the meantime, marketing departments investing in retail media should require each platform to provide a documented measurement methodology. Three criteria help evaluate the reliability of retail media reporting:
- The distinction between attributed sales (the consumer saw the advertisement and then purchased) and incremental sales (they would not have purchased without the advertisement), along with the method used to separate them.
- The ability to compare indicators with those from other channels (search, social) on a comparable basis, which requires shared definitions of cost per acquisition or return on investment.
- Access to raw or semi-aggregated data, rather than a proprietary dashboard where calculations cannot be verified.
Without these elements, there is a risk of reallocating budget to a channel whose actual contribution is poorly measured.
Visibility in AI Response Engines: A Marketing Indicator in Its Own Right
When a user asks a question to ChatGPT, Perplexity, or Google’s AI Overviews, the response sometimes cites brands, products, or sources. Being mentioned in an AI response becomes a visibility issue distinct from traditional SEO.
The problem is that this visibility largely escapes the usual dashboards. Referral traffic from an AI response does not always clearly appear in Google Analytics. Some publishers find that their content feeds generated responses without a click to their site materializing.
For marketing teams, this means tracking not only incoming traffic but also the presence and citations of the brand in generated responses. There is talk of GEO (Generative Engine Optimization) to designate this discipline. In practical terms, this involves:
- Structuring pages with clear answers to specific questions (FAQ format, short paragraphs with verifiable information per sentence).
- Multiplying trust signals (customer reviews, mentions by experts, sourced data) that language models identify as reliable.
- Regularly monitoring whether the brand appears in AI responses for its strategic queries, manually or via emerging dedicated tools.
GEO does not replace SEO, but both deserve separate tracking. A brand well-positioned on Google can be completely absent from generative responses, and vice versa.
First-Party Data and Personalization: Balancing Depth and Volume
With the gradual disappearance of third-party cookies, first-party data becomes the main fuel for marketing personalization. Brands capable of collecting information directly from their audiences (customer accounts, forms, loyalty programs) have a concrete operational advantage.
The temptation is to multiply collection points to inflate the database. Collecting fewer but more actionable data yields better results than a poorly structured large database. A well-thought-out three-field form, which immediately segments the profile (need, budget, deadline), feeds more effective automated marketing scenarios than a database of ten thousand unqualified contacts.
The shift to first-party data also alters the relationship between marketing and IT. Personalized email scenarios, dynamic product recommendations, or adaptive site journeys rely on a data infrastructure that the marketing team cannot manage alone. The ability to collaborate between the two functions on a common data framework determines the quality of personalization as much as the choice of tool.
Results vary significantly by sector and the digital maturity of the company. A B2B SME with a long sales cycle will not derive the same benefit from a hyper-personalization scenario as a low-average basket public e-commerce merchant. The balance between the depth of personalization and the resources mobilized remains the real operational issue.
Modern marketing is not just about adopting new tools. Each lever (generative AI, retail media, GEO, proprietary data) comes with measurement, compliance, or organizational constraints that determine its actual effectiveness. Teams that progress are those that treat these constraints as integral to their strategy, not as obstacles to be circumvented.
