Participate in the drawing-up of the Code of Practice on transparent generative AI systems

The AI Office invites AI providers of generative AI systems, providers of transparency techniques, associations of deployers of in-scope AI systems, civil society organisations, academic experts, and other relevant organisations, to express their interest to participate in the drawing-up of the Code of Practice. 

This Code of Practice will outline a way for providers and deployers of generative AI systems to demonstrate compliance with the transparency obligations of Article 50 AI Act which will become effective on 2 August 2026. AI providers must ensure that artificially generated or manipulated content is detectable through machine readable marks. Deployers will also have to disclose that the content has been artificially generated or manipulated in the case of deep fakes or text published with the purpose of informing the public on matters of public interest. These obligations aim to reduce the risks of deception, manipulation and misinformation.

The drafting process will begin in October 2025 and is expected to take about 10 months. The approach, facilitated by the AI Office, will be inclusive, transparent and iterative and involve AI independent experts.

Two thematic working groups will be led by Chairs and Vice-Chairs appointed by the European AI Office, with the participation of interested eligible stakeholders. The Chairs will prepare the first iteration of the draft, taking into account the results of a parallel consultation launched on 4 September 2025.

Expressions of interest shall be submitted by Thursday, 2 October 2025, 23:59 CET via this application form.