From 2 August 2026, the transparency obligations under Article 50 of the Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on Artificial Intelligence (AI Act) apply. These rules require:
- individuals to be informed when they are interacting with AI or are exposed to emotion recognition or biometric categorisation systems;
- AI-generated or AI-manipulated content to be identifiable through technical measures;
- certain content (deepfakes and AI-generated or AI-manipulated text on matters of public interest) to be clearly labelled as AI generated or manipulated.
The four new obligations
Article 50 of the AI Act introduces four distinct transparency obligations, which apply either to providers or to deployers of AI systems, depending on the role an organisation performs in relation to a particular use case. Importantly, the same organisation may perform both roles in different circumstances, depending how it is used.
1 Disclosure of AI-human interactions – Article 50(1)
Providers of AI systems intended to interact directly with natural persons (e.g. chatbots or voicebots) must ensure that individuals are informed that they are interacting with an AI system.
2 Machine-readable marking of AI-generated content – Article 50(2)
Providers of AI systems, including general-purpose AI models, that generate synthetic text, images, audio or video must ensure that AI-generated outputs are marked in a machine-readable format and are detectable as artificially generated or manipulated.
3 Transparency for emotion recognition and biometric categorisation systems – Article 50(3)
Deployers using emotion recognition or biometric categorisation systems must inform individuals exposed to those systems about their operation, no later than at the time of their first exposure.
4 Disclosure of AI-generated or manipulated content (deep fakes and certain AI-generated text) – Article 50(4)
Deployers using AI systems to generate or manipulate image, audio or video content constituting a deep fake must clearly disclose that the content has been artificially generated or manipulated.
Article 50(1) of the AI Act will be particularly relevant for companies that use chatbots or voicebots to communicate with customers. For companies that do not develop AI systems themselves but use commercially available AI tools (e.g. popular generative AI tools such as DALL-E, Midjourney or Stable Diffusion), for example to create marketing content, Article 50(4) will be of particular importance.
Practical challenges
To achieve compliance with the requirements of Article 50 of the AI Act, several key issues need to be addressed. First, a company must correctly determine its role in the process of creating and publishing content and, consequently, identify the obligations that apply to it. For example, different requirements apply to the provider of an image-generation tool and to a company that uses AI-generated material in advertising or publishes it on social media.
Second, the appropriate method of labelling AI-generated content must be selected depending on the company’s role in the process and the type of content involved. In some cases, the relevant disclosure may take the form of metadata that is not visible to the user, while in others – for example, when publishing a deepfake – a visible disclosure informing the audience that AI has been used may be required.
Finally, a key challenge is determining whether particular content constitutes a deepfake within the meaning of the AI Act. For example, would this include an AI-generated image of a fictional model, a digitally altered face of a real person, or merely an AI-assisted retouching of a photograph?
Next steps
We recommend conducting an internal assessment to identify which AI systems are used within the organisation, determine the organisation’s role in the AI supply chain, and assess which transparency obligations apply.
We also recommend reviewing:
- information notices provided to customers when interacting with AI;
- internal processes for labelling AI-generated content;
- internal policies governing employees’ use of AI tools.
If you have any questions regarding the transparency obligations under the AI Act, please get in touch.
