Europe’s new deepfake label has a private-use blind spot

Europe’s new deepfake rules require a visible warning from many users of AI — but not when a private individual makes and shares the fake for personal, non-professional use.

Article 50 of the EU AI Act began applying in Sweden on 2 August. Professional deployers must disclose deepfake image, audio or video clearly when people first encounter it. The European Commission says an embedded machine-readable marker alone does not satisfy that disclosure duty.

The legal line turns on who is using the system. A person generating and posting a deepfake privately is outside the Act’s deployer rules. If the activity regularly earns money or forms part of a business, trade, job or freelance work, it becomes professional and the duty applies.

AI-system providers still have a separate obligation to make generated or manipulated output machine-readable and detectable. That safeguard is not the same as a warning ordinary viewers can necessarily see.

The practical consequence for people scrolling in Sweden is simple: an unlabelled clip is not proof that it is authentic. The new rule makes professional deception harder to hide, but it does not make every private account announce what AI changed.

DailySweden

DailySweden's desk of editors, reporters, researchers and investigative journalists. We bring readers factual, truthful and objective reporting on the issues that matter to them.

Contact the editorial desk ›