In early September 2025, the European Commission launched two linked initiatives:

• a public consultation to draft guidelines on transparency obligations under the EU AI Act, and
• a call for expression of interest for stakeholders to help co-design the first Code of Practice on Transparent Generative AI Systems.

These initiatives aim to translate Article 50 transparency rules of the AI Act into practical measures that providers and deployers of AI can apply in day-to-day products and services.
 

read the original article here

What it is
• Guidelines (under preparation) - to clarify transparency requirements: who must disclose what, when, and how, including exceptions and sectoral considerations.
• Code of Practice (voluntary, co-designed with industry and civil society) to develop concrete tools and standards for:
1. marking and labeling AI-generated or manipulated content (including machine-readable formats),
2. making AI outputs reliably detectable,
3. reporting provenance and authenticity information across the value chain,
4. ensuring clear and accessible end-user disclosures when people interact with AI systems.

Why it matters
• Operational clarity: turns broad legal duties in the AI Act into actionable compliance measures for companies.
• Trust and oversight: makes AI-generated or manipulated outputs easier to identify and audit, a foundation for user trust, enforcement, and meaningful human oversight.
• Risk reduction: reduces legal and reputational risks by clarifying expectations for providers and deployers before Article 50 obligations become binding in August 2026.
• Ecosystem alignment: supports convergence of technical approaches (labeling, detection, provenance) so that businesses, regulators, and users can verify compliance consistently.

Timeline
• Consultation & expression of interest open until 2 October 2025.
• Drafting of the Code of Practice will start in October 2025, with work expected to run for about 10 months.
• Transparency obligations in Article 50 of the AI Act apply from August 2026.

The EU is treating transparency as the cornerstone of safe, trustworthy AI. These initiatives are not just about compliance, they are also about creating common technical and procedural standards that can actually work in practice across Europe.

Navigating the complexities of AI Transparency can be challenging. If your company needs support in understanding and applying these new rules, we’d be happy to connect and explore how we can assist.