A ACT AI NOW
A framework by ACT AI NOW · Jeannette Gorzala

Disclosed. Labelled. Traceable.

Article 50 of the EU AI Act requires AI transparency in interactions and content. CLEAR is the governance framework that turns those obligations into an operational system — so AI interactions are disclosed, AI-generated content is labelled, and both become traceable and evidenced across your organisation, consistently rather than case by case.

Article 50 of the EU AI Act · in force 2 August 2026.

The problem

AI is scaling faster than the systems meant to keep it transparent.

AI is now deeply embedded in how organisations communicate, create content and make decisions. The obligation to be transparent about it has become reality. Most organisations can confirm that they use AI applications, but cannot show where exactly it is used in their content, label it consistently, or make its use traceable on request.

AI is now embedded in

  • — Communication and marketing systems
  • — Content production pipelines
  • — Decision and approval workflows
  • — Customer-facing conversational interfaces

But transparency stays fragmented

  • — Labelling, without understanding the system behind it
  • — Compliance, without visibility into the architecture
  • — Governance, without traceability
  • — Documentation that can't reconstruct what happened

You can see that AI is being used. What usually can't be traced is which content it generated, whether that was disclosed, and where accountability sits. That is exactly the gap regulators, supervisory bodies or courts will confront you with.

Why now

Media isn't a seamless value-chain. It's an unsynchronised system.

With the EU AI Act's transparency obligations now in force, organisations must not only know the rules but implement them and operate in compliance. The implementation problem is structural. The media value chain consists of many different actors: model developers, AI system providers, AI system deployers, distributors and audiences. In many instances, content is modified with and without AI applications, without traceability or documentation.

In force

2 Aug 2026

Labelling & visual transparency

Penalty

€15M / 3%

Maximum exposure

In force

2 Dec 2026

Embedded & technical transparency

RIGHTS & KNOW-HOW BACKFLOW behaviour & re-shares → next-gen training signal ORIGINATE Creators & sources photographers writers / journalists musicians · illustrators archives · UGC BUILD Model developers foundation models fine-tuners / open-weight ASSEMBLE System providers APIs & products tooling / infra DEPLOY Deployers enterprises public sector agencies professionals AMPLIFY Distributors & platforms social platforms aggregators / search ad networks news agencies (wholesale) CONSUME Consumers & public citizens · customers audiences MEDIA PRODUCERS — a subset of deployers under the AI Act journalists · press agencies · broadcasters · publishers · influencers · marketers
The media system: content flows forward; rights and training signal flow back.

Many participants, a heterogeneous ecosystem, no content provenance. When regulation asks who created something, with what, and whether it was disclosed, the value chain often cannot answer. That is exactly the gap CLEAR exists to close — by anchoring transparency systemically instead of leaving it to individual improvisation.

The reference document

The CLEAR Whitepaper

Why thinking about transparency only at the output stage stops working once AI sits inside your systems — and how CLEAR maps the EU AI Act obligations to the architecture that has to evidence it.

Explore the framework
A Whitepaper · 2026

Transparency
by Design

A CLEAR framework for AI transparency under the EU AI Act, Article 50.

ACT AI NOWJeannette Gorzala

The framework

Five dimensions of AI transparency.

CLEAR isn't a checklist — it's a model for systematically designing transparency into the system. The framework defines the operationalization; the instruments guide implementation into practice.

C

Conversational Transparency

Interaction transparency. People are informed that they are interacting with an AI system — throughout the whole interaction, not just at the first screen.

L

Labelling

Media transparency. AI-generated and AI-edited content is visually labelled as required by the EU AI Act. Implementation is standardised across images, video, audio and text.

E

Embedded Watermarking

Synthetic content is marked technically and machine-readably. This is embedded into the workflow and architecture — by design, not after the fact.

A

Accountability

Clear responsibilities across the supply chain and for internal teams and processes are defined and agreed.

R

Regulatory Readiness

The wider applicable regulatory requirements (e.g. GDPR, copyright) are accounted for — including evidence on request.

CLEAR is not a checklist. It is the systematic implementation of Transparency by Design.

The CLEAR Score

How transparent is your AI approach, really?

Ten questions across the five CLEAR dimensions. You'll get a maturity assessment, a breakdown by dimension and a gap assessment. For orientation.

Scale: Not at all · Partially · Largely · Fully

Implementation

Put CLEAR into practice —
with the working artefacts.

The CLEAR working artefacts assist you in turning the framework into the transparency and governance standards your team runs day to day. We can share our standard best practices or customize together.

AI labelling guideline

AI content labelling icons

Governance structure template

Accountability mapping

Regulatory readiness review

Advisory

We design and implement the system with you.

Led by Jeannette Gorzala — who helped shape the AI Act and now advises on how it is applied — CLEAR Advisory takes you from assessment to a transparency system that holds up.

Services

  • — Governance design
  • — System & transparency audits
  • — EU AI Act readiness
  • — Transparency architecture

Formats

  • — Executive & board advisory
  • — Workshops
  • — Implementation programmes
  • — Ongoing oversight support

Start with CLEAR

Know where you stand. Then build the system.

Tell us what you're working on and we'll come back to you. Or calculate your CLEAR Score first to see where you stand.

Take the CLEAR Score →