Starting August 2, 2026, transparency rules under the AI Act will apply in the European Union. In certain cases, companies, media outlets, and other professional users must clearly indicate that published text, images, audio, or video were created or substantially modified using artificial intelligence. However, this requirement does not apply to every minor modification or to purely private use of AI.
The strictest rules apply to so-called “deepfake” content—that is, realistic images, videos, or audio recordings that may appear authentic. The label may take the form of text, an icon, a watermark, or an audio alert. It must be clear, understandable, and recognizable no later than the first time the content is displayed or played.
Special rules also apply to texts intended to inform the public about matters of public interest. If such an article was created or significantly edited using AI, its use must be disclosed. However, labeling is not required if the text has undergone genuine human or editorial review and a specific person or publisher has assumed responsibility for its publication. Routine spelling corrections, color filters, or minor cosmetic adjustments to photographs therefore generally do not, in and of themselves, give rise to a labeling requirement.
Our team of experienced attorneys will help you solve any legal issue. Within 24 hours we’ll evaluate your situation and suggest a step-by-step solution, including all costs. The price for this proposal is only CZK 690, and this is refunded to you when you order service from us.