Communication & media · AI Act

Communication & media — AI Act: content transparency

The AI Act imposes transparency obligations directly relevant to content production: labelling AI-generated or AI-manipulated content, flagging deepfakes, and informing users when they are interacting with an AI. This guide details these obligations and the applicable exceptions.

DPO / CISO team — Data Comply One Updated on 15 June 2026

In brief

  • Regulation: AI Act.
  • Agencies, media outlets, ad networks, and creators deploying generative AI systems are concerned as deployers; tool publishers are providers.
  • Watermark AI-generated content.
  • Report deepfakes.
  • Informing about AI interaction.
  • Up to €35M or 7% for prohibited practices and €15M or 3% for other serious infringements.

Regulatory deadlines

The key dates of this regulation.

August 2024

Entry into force

In force

February 2025

Ban on prohibited practices

In force

August 2025

General-purpose AI models (GPAI) & governance

In force

August 2026

High-risk system obligations (Annex III)

In force

August 2027

High-risk embedded in products (Annex I)

Upcoming

What does the AI Act cover?

The AI Act establishes transparency obligations: providers of content-generating systems must enable their output to be labelled, and deployers must flag manipulated content (deepfakes), with exceptions (e.g. satirical or artistic works). Interaction with an AI system (chatbots) must be disclosed.

Manipulative practices exploiting vulnerabilities are prohibited.

Is the 'Communication & media' sector affected?

Agencies, media outlets, ad networks, and creators deploying generative AI systems are concerned as deployers; tool publishers are providers.

Transparency applies regardless of the medium (text, image, voice, video).

Detailed obligations

Label AI-generated content

Disclosing content generated or substantially manipulated by AI.

Report deepfakes

Clearly indicate manipulated content depicting real people or events, outside permitted exceptions.

Inform on AI interaction

Notify when a user interacts with an AI.

Avoid manipulation

Ban prohibited influence techniques.

Articulate with GDPR

Comply with GDPR and portrait rights for content featuring individuals.

Sanctions & risks

Up to €35M or 7% for prohibited practices and €15M or 3% for other serious infringements. An unlabelled deepfake combines AI Act, GDPR, and image rights risks.

Application timeline

  • 1Entry into force. 1 August 2024.
  • 2Prohibitions. Since February 2025.
  • 3Transparency. Progressively applicable.

Common mistakes in the sector

  • 1Unmarked content. Distributing AI-generated content without transparency.
  • 2Unreported deepfake. Publishing a deepfake without disclosure.
  • 3Manipulation. Using prohibited influence techniques.
  • 4Image not respected. Overlooking portrait rights and GDPR.

Practical case

An agency produces a promotional video using a synthetic voice imitating a real person. It discloses the generated nature of the content, obtains the necessary authorisations under image rights and GDPR, and avoids any misleading presentation.

Compliance roadmap

  1. 1

    Inventory. Map generative AI use cases.

  2. 2

    Mark. Implement content labelling.

  3. 3

    Report. Report deepfakes and AI interactions.

  4. 4

    Frame. Ban manipulative practices.

  5. 5

    Articulate. Comply with GDPR and image rights.

Frequently asked questions

Content that is generated or substantially manipulated must be labelled as such; exceptions exist (artistic, satirical).

Back to the Communication & media guideCommunication & media & GDPR

Take action on AI Act

Free assessment or a chat with an expert dedicated to communication & media.

Start my free assessment Talk to an expert

They already trust us

See how organisations in the communication & media sector secured their compliance with DCO.

See testimonials