Earlier this week, the European Commission published finalised Guidelines on Transparency of AI-Generated Content. These clarify the obligations under Article 50 of the EU AI Act, which apply from 2 August 2026. The Oireachtas has also approved Ireland’s implementing legislation, with Ireland’s new AI office to be established by the same date.
A policy briefing is available here for members setting out what the guidelines cover, what they mean for the Creative & Communications sector, and the steps available to demonstrate compliance. Also available is a selection of the European Commission’s FAQs, with relevant examples.
Key points to note:
- Emotion recognition and biometric categorisation tools: If a campaign or platform uses AI to analyse audience emotions or categorise people biometrically, individuals must be informed.
- Deepfakes: any AI-generated or manipulated image, audio or video resembling real persons, objects or places that could falsely appear authentic must be disclosed to the audience upon first exposure, via clear and perceivable labels (visible or audible). Machine-readable marks alone are not sufficient. Note that this is context-sensitive — if the intended audience clearly does not expect authenticity (e.g. an obviously fantastical effect), it may not meet the definition.
- AI-generated text on matters of public interest applies if AI-generated copy on political, economic or social topics is published without human review or editorial control.
Whilst the legislation will be in effect from 2 August, it is expected that the initial focus will be on establishing the AI office, which will then have responsibility for coordinating the designation of each regulator as a competent authority. No retrospective labelling will be required.
For more information, contact Niall Carty, Director of Public Affairs: niall.carty@ccai.ie
