EU AI Act Transparency Rules Take Effect, Requiring Disclosure of AI Interactions and Synthetic Content
Transparency obligations under the European Union’s AI Act began applying on August 2, 2026, creating a new legal baseline for how certain artificial-intelligence systems and synthetic media are identified across the bloc.
The European Commission says the requirements, based on Article 50 of the AI Act, cover disclosures to people interacting directly with AI and markings for certain AI-generated or manipulated content. The rules affect covered AI providers, deployers and online platforms operating within the European Union.
What changed on August 2
Providers must inform people when they are directly interacting with an AI system in circumstances covered by the law. That requirement is aimed at making the nature of the interaction clear rather than allowing an AI system to present itself as a human without disclosure where the obligation applies.
The rules also require certain AI-generated or manipulated content to carry machine-readable markings. These markings are intended to identify covered synthetic material in a way that can be processed by technical systems, although the Commission’s materials do not say that every piece of AI-generated content must be labeled in every circumstance.
The obligations vary according to the type of system and the applicable legal exemptions. That means the start date is a broad compliance milestone, not a declaration that all AI tools, images, video, audio or text fall under identical requirements.
Deepfakes and public-interest content
Deployers have additional disclosure duties for deepfakes and for certain AI-generated public-interest content created without human editorial review. The distinction matters because the obligations are not limited to the companies that build or supply AI systems. Organizations that deploy covered systems or publish covered material may also have responsibilities.
For readers, the practical effect is a new set of disclosure expectations around some interactive AI services and synthetic media encountered through platforms or other public-facing systems in the EU. The requirements are designed to address confusion about whether a person is communicating with AI and whether media has been generated or manipulated.
The rules do not, by themselves, establish that labeled content is authentic, accurate or safe. A marking can identify how content was produced or altered, but it does not settle the truth of the content’s claims or guarantee that a system complied correctly.
How the compliance process reached this point
The Commission opened a consultation on draft guidance for the transparency obligations on May 6, 2026. In that material, it described Article 50 as the legal basis and outlined obligations for providers and deployers as well as the planned compliance timeline.
On July 20, the Commission published guidance explaining the obligations. The guidance addressed disclosure of AI interactions, marking of generated or manipulated content, and covered categories including deepfakes and certain public-interest AI content.
The August 2 application date therefore marks the transition from preparation and draft guidance to the point at which the transparency obligations began applying. It is a final application date for the obligations described in the approved Commission materials, not merely a proposal or consultation deadline.
What happens next
Providers and deployers must now assess which of their systems and outputs fall within the covered categories, apply the relevant disclosures or markings, and account for the legal exemptions and system-specific scope described by the Commission.
The approved sources do not detail the first enforcement actions or penalties after the rules took effect. They also do not establish how individual platforms or providers are complying in practice. Further enforcement activity, guidance or case-specific decisions would be needed to show how the requirements are applied in particular situations.
For now, the material change is the effective date itself: as of August 2, 2026, Article 50 transparency obligations apply across the European Union to the covered AI interactions and synthetic-content categories identified by the Commission.
Sources
Discover more from Interactive News
Subscribe to get the latest posts sent to your email.
