On August 2, 2026, part of Regulation (EU) 2024/1689 of the European Parliament and of the Council of June 13, 2024, establishing harmonized rules on artificial intelligence (the AI Act) entered into force. Among other provisions, Article 50 of the AI Act, concerning the disclosure of content generated or manipulated using artificial intelligence, is now in effect.

The AI Act does not impose an obligation to disclose all content created using AI. This obligation applies primarily to so-called “deepfakes,” as defined in Article 3(60) of the Regulation. A deepfake is:

1. content generated or manipulated by AI;
2. in the form of image, audio, or video content;
3. resembling existing persons, objects, places, entities, or events;
4. which would falsely appear to a person to be authentic or truthful.

If all these conditions are met, there arises an obligation to appropriately disclose that the content has been generated or manipulated using artificial intelligence. However, the Guidelines on transparency obligations for providers and deployers of AI systems provide an exception from the obligation when AI-supported manipulation of insignificant substantive or technical aspects of pre-existing content may be of minor relevance for a person’s assessment of the authenticity or truthfulness of the content, not rendering the resulting content to become a deep fake.

The obligation to disclose content created using AI applies to all deployers of AI systems. The Regulation requires that information regarding the use of artificial intelligence be disclosed to the recipient in a clear and distinguishable manner, at the latest at the time of the first interaction or exposure. The European Commission has developed a Code of Practice on the Transparency of AI-Generated Content, which provides a set of icons that may be used to fulfill the obligations under Article 50 of the Regulation. However, deployers may also use equivalent disclosures.

Under Article 50 of the AI Act, the obligation to disclose content generated or manipulated using AI also applies to deployers publishing AI-generated or manipulated text informing the public on matters of public interest, provided that such content has not undergone human review or editorial control and that no natural or legal person holds editorial responsibility for the publication of the content.

The Commission has also published Guidelines on the transparency obligations of providers and entities implementing artificial intelligence systems, which specify an exception for insignificant changes made by AI to technical or substantive aspects. Content modified to this extent is not subject to the disclosure requirement.

The obligations described above apply to content created on or after August 2, 2026. There is no obligation to disclose AI generated or manipulated content created prior to that date (even if unpublished). However, it is recommended to retain evidence proving that the content was created before August 2, 2026.

Until a practice is established allowing the use of the EU icons “AI GENERATED” and “AI MODIFIED,” in Poland it is recommended to use their equivalents in Polish, i.e., “WYGENEROWANE PRZEZ AI” or “ZMODYFIKOWANE PRZEZ AI.” This stems from the provisions of the Polish Language Act of October 7^th, 1999, which mandates the use of the Polish language in communications with consumers; this also applies to warnings and consumer information required under other regulations.

In the coming months, further guidelines from the European Commission, positions from supervisory authorities, and the first decisions and rulings clarifying how to apply the new regulations are likely to emerge. The Polish Act on Artificial Intelligence Systems of July 3^rd, 2026, which establishes a national supervisory authority for the AI Act, will also be significant for the practical application of these regulations.