A missing AI notice does not make content genuine. This check explains when responsible users, companies and organisations must disclose AI-generated or AI-manipulated content. It does not determine whether content is genuine or fake. Anyone intending to deceive or defraud may deliberately ignore a disclosure obligation. The absence of an AI notice therefore never proves that an image, video, audio recording or text is authentic.

Covered are the information and disclosure obligations of deployers under Article 50(3) and (4) – deepfakes, published AI texts, emotion recognition and biometric categorisation. The notification duty for chatbots under paragraph 1 is covered as well.

Not fully covered are the technical provider obligations under Article 50(2), that is the machine-readable marking of generative systems. Nor are special cases of legally authorised law enforcement.

On the frequently cited date of 2 December 2026: it applies only to the machine-readable provider marking under Article 50(2) for generative systems already placed on the market before 2 August 2026. The information and disclosure obligations under Article 50(1), (3) and (4) have applied since 2 August. The claim that all disclosure obligations have been postponed until December is not correct.