New European Union rules will mandate that artificially generated images, audio, and text designed to look authentic must be labelled. These regulations, part of the EU’s AI Act, come into effect for new AI systems on the EU market from Sunday, 2 August. Existing systems will have an extra four months to meet the requirements.
The rules aim to ensure that people are aware when they are interacting with AI-powered chatbots or viewing manipulated images or AI-generated text. Synthetic content intended to appear truthful must be visibly marked as AI-generated and include a digital watermark indicating its artificial origins. Texts on matters of public interest also require labelling if there is no human editorial oversight.
Green MEP Sergey Lagodinsky, who helped negotiate the AI Act, stated that making this information transparent is crucial for preserving democracy and the authenticity of facts online. Companies that fail to comply with the regulations could face fines of up to €15 million or 3% of their worldwide global turnover.
The tech industry has expressed support for making deepfakes visible but has raised concerns that the rules might be interpreted too broadly, potentially overwhelming internet users with labels. Boniface de Champris, CCIA Europe’s AI policy lead, suggested that the most noticeable change for the public might be in areas like advertising, film, and publishing, rather than social media.