Workshops: EU AI Act & Compliance

Do You Need to Label AI Content? Transparency Workshop

The EU AI Act requires companies to disclose AI in certain situations, for instance when people interact with a chatbot or when images, video or audio have been realistically generated. In this workshop you clarify which of your channels are affected and set concrete labelling rules.

Who is this workshop for?

  • Marketing, content and social media
  • Customer service teams and chatbot owners
  • Product managers of digital services
  • Corporate communications
  • Compliance and data protection

What you will take away

  • Know the AI Act transparency duties for chatbots, synthetic media, deepfakes and certain AI-generated text.
  • Distinguish when the duty lies with the tool provider and when with you as deployer.
  • Understand how technical marking such as watermarks and metadata works and where it falls short.
  • Agree on concrete wording and placement of AI notices for your channels.
  • Give your team a simple decision aid: label or not.

Agenda

  1. Module 1

    Transparency in the EU AI Act

    • Provider and deployer duties at a glance
    • Chatbots and AI interaction
    • Synthetic media, deepfakes and text on matters of public interest
  2. Module 2

    Reviewing your channels

    • Exercise: website chat, social media, newsletters, product images, videos
    • Exceptions such as editorial control or artistic works
    • Relation to advertising law and platform rules
  3. Module 3

    Implementing labels

    • Visible notices: wording and placement
    • Technical marking: metadata, content credentials, watermarks
    • What ChatGPT, Claude, Google Gemini or Microsoft 365 Copilot already provide
  4. Module 4

    Process and decision aid

    • Decision tree for the content team
    • Approval checklist
    • Documentation as evidence

What the transparency duties are about

People should be able to tell when they are interacting with AI or when realistic-looking content was artificially generated. The AI Act splits duties between providers, who should enable machine-readable marking, and deployers, who must for example disclose deepfakes.

Not every AI-assisted text needs a label. That is exactly why teams need clear criteria instead of labelling everything or nothing.

More than ticking a box

Transparency is also about trust. Customers react strongly when they feel misled. Consistent, well-considered labelling can strengthen your brand.

  • Consistent language across all channels
  • Clear rules for agencies and freelancers
  • Less uncertainty in day-to-day publishing

Please note

The details of the transparency duties are being refined through guidelines and codes of practice. We cover the current state and build flexible processes; the workshop is not legal advice.

Frequently asked questions

Does every AI-assisted blog post need a label?

Not across the board. The duties mainly target realistic synthetic media and certain texts on matters of public interest, where editorial control plays a role. We show you how to classify your content.

Is a notice in the legal imprint enough?

Usually not. Notices should be clearly visible at the latest at first interaction or exposure. In the workshop we develop suitable placements.

What applies to our customer service chatbot?

Users generally need to be told they are interacting with AI unless it is obvious. We draft a suitable notice for your bot together.

Workshop

Get in touch

Want clear labelling rules across all your channels?

Tell us a little about your team – we will get back to you within 24 hours with a proposal for content, format and dates.

Prefer to talk?

We only use your details to handle this request. See our privacy policy.