However, this does not mean that there is a general obligation to label every piece of content created using AI. This obligation applies only to the cases specified in that provision. Therefore, the manner in which a given AI system is used must be assessed on a case-by-case basis. It is advisable to conduct such an assessment now and prepare internal procedures accordingly.

What Has Changed Since August 2, 2026

In the EU, regulations governing the use of artificial intelligence have been in effect since August 1, 2024, under the AI Act. The timeline for the application of some of the most demanding obligations under the AI Act was amended by Regulation (EU) 2026/1744 of the European Parliament and of the Council of July 8, 2026. Consequently, as of August 2, 2026, the new transparency requirements set forth in Article 50 of the AI Act are in effect.

Key Obligations

The new transparency obligations include:

✓ Disclosure of interactions with chatbots and AI: Providers must ensure that users are informed when they are interacting with an AI-based system, unless this is clearly evident from the context.

✓ Labelling of synthetic content: Providers must ensure that synthetic audio, image, video or text content is marked in a machine-readable format and detectable as artificially generated, subject to certain exceptions.

✓ Labelling of deepfake content: Entities using artificial intelligence to create deepfake content must clearly disclose that the content has been artificially generated or manipulated.

✓ Notifications regarding emotion recognition and biometric categorization: Entities implementing these systems must inform the individuals who interact with them.

Depending on whether you are a provider (for example, by developing a chatbot) and/or a user (for example, by posting AI-generated content on a website), different labelling requirements may apply. Therefore, as a first step, your company should assess its specific role and determine whether you are a provider and/or a user.

Penalties

Penalties for non-compliance with the new obligations depend on the severity of the violation and the size of the company concerned. They may amount to up to 15 million euros or up to 3 percent of the company’s total global turnover for the preceding financial year.

Actions

It is worth using the additional time before the next phases of the AI Act take effect to streamline processes, conduct a compliance audit, and implement internal procedures. The recommended actions include:

• Identify and classify the AI systems used by the company: It is advisable to map the solutions currently in use by the company.
• Review transparency obligations (Article 50 of the AI Act): This review should help determine whether your organization uses artificial intelligence in a manner that requires disclosure and transparency. This includes, among other things, synthetic content, deepfakes, and materials that could mislead recipients as to their origin.
• Develop internal policies: It is advisable to establish internal procedures governing the safe and lawful use of AI.
• Review contracts with AI tool providers: in order to clarify the division of responsibility for compliance with EU regulations.

•    Employee training.