Media companies develop AI transparency guidelines

Dutch and Belgian media companies are collaborating to establish transparency guidelines for the use of artificial intelligence in image, audio, and video content. Initiated by Media Campus NL, the project includes major organizations such as NPO, DPG Media, Mediahuis, and Talpa.

The development of these standards is driven by the European AI Act, which mandates that media organizations clearly disclose the use of AI in specific instances, such as deepfakes. The guidelines aim to assist creators in determining when content qualifies as a deepfake, when an AI label is required, and the appropriate duration for such disclosures.

Different standards will apply depending on the content type, with news and journalism subject to different principles than entertainment, reality, or fiction. Technical applications that do not alter the content, meaning, or context will generally not require a label. Guidelines for AI-generated text are also expected to be released soon.