EU AI labels become mandatory from Aug 2

Deepfakes and other AI-generated content must be labelled in the European Union from Aug 2 as the bloc begins enforcing transparency provisions in its AI law. Companies must make clear when users are interacting with AI systems such as chatbots, and content such as images or text created using AI must carry labels, watermarks or other markers that make synthetic media easier to detect.

The requirements cover professional content and text intended to inform the public on general-interest issues when it is created using AI without human editorial oversight. Individuals using AI purely for personal purposes are not covered. Existing AI systems have until Dec 2 to adapt, and exemptions apply to “artistic, creative, satirical (and) fictional” work.

Large technology companies have already moved toward compliance. TikTok says more than three billion pieces of content have labels through creator disclosures and detection tools, Meta uses an “AI Info” label on Instagram and Facebook, and Google has signed the EU code of conduct while working with Nvidia, OpenAI and Apple on digital tagging. Critics warn the added regulatory complexity could produce overlapping disclosures that make labels less useful for users.