How to implement DSA ad transparency and recommender transparency as real UX and data controls.
Covers platform and VLOP/VLOSE layers: per-ad disclosures, recommender parameter transparency, ad repositories, and non-profiling options.
Structured answer sets in this page tree.
Cited legal and guidance references.
DSA advertising and recommender obligations are user-facing controls. Compliance isn't a policy statement - it's UI behavior, logging, and the ability to prove what a user saw and why. This page translates Articles 25-28 and (for VLOPs/VLOSEs) Articles 38-39 into implementable requirements, UX patterns, and evidence expectations.
The DSA adds transparency and user control where ranking and targeting influence what people see.
The obligations focus on: clear labeling, attribution (who benefits and who paid), meaningful targeting explanations, and user control over recommender parameters.
If you present advertisements on your online interface, Article 26 requires that for each specific ad and each recipient, they can identify certain information clearly, concisely, unambiguously and in real time.
Treat this as a UI requirement plus an instrumentation requirement.
If you use recommender systems, Article 27 requires you to explain the main parameters in your terms and conditions, and describe options for recipients to modify or influence those parameters.
In practice, this means "why am I seeing this?" plus a control surface (where options exist).
Article 25 prohibits designing/organising/operating interfaces in a way that deceives or manipulates recipients or materially impairs their ability to make free and informed decisions.
Ads and recommender controls are common dark-pattern surfaces (consent, personalization toggles, cancellation flows).
If your platform is accessible to minors, Article 28 requires appropriate and proportionate measures to ensure a high level of privacy, safety, and security of minors on the service.
Article 28 also restricts presenting ads to minors based on profiling when you are aware with reasonable certainty the recipient is a minor, without forcing additional personal data processing solely to assess age.
For VLOPs/VLOSEs that use recommender systems, Article 38 adds a specific requirement: provide at least one option for each recommender system that is not based on profiling (GDPR profiling definition referenced).
This is a major product requirement: you need an alternative ranking mode and a control surface.
Article 39 requires VLOPs/VLOSEs presenting ads to make a public repository available via a searchable tool (multi-criteria queries) and via APIs, covering the period the ad is shown and for one year after it was last shown.
Build this as a product and data platform capability, with privacy protections.
Ads and recommender compliance is often challenged by "prove it" questions: what did the user see, what did it mean, and what options existed?
Build evidence and testing into your release process.
Research Copilot can take EU Digital Services Act (DSA) Ads & Recommender Systems from getting cited answers and faster research on this topic to a reusable workflow inside Sorena. Teams working on EU Digital Services Act (DSA) can keep owners, evidence, and next steps aligned without copying this guide into separate documents.
Start from EU Digital Services Act (DSA) Ads & Recommender Systems and answer scope, timing, and interpretation questions with cited outputs.
Review your current process, evidence gaps, and next steps for EU Digital Services Act (DSA) Ads & Recommender Systems.