On 2 August, new transparency obligations became applicable to providers and deployers of certain Artificial Intelligence systems.
Among the new obligations, the following stand out:
1. Identification and labelling of AI-generated content
Deepfakes, emotion recognition or biometric categorisation systems, as well as certain text-based content relating to matters of public interest, where generated or manipulated by AI, must be appropriately identified.
Such content must also contain machine-readable marks enabling it to be recognised as having been generated or manipulated through Artificial Intelligence.
2. Transparency when interacting with AI systems
Users must be clearly informed when they are interacting with an AI system — rather than with a person.
This is the case, for example, when interacting with chatbots, AI agents or avatars.
The European Commission has published guidelines aimed at supporting compliance with these obligations, highlighting the Code of Conduct on the Transparency of AI-Generated Content as a tool to support the demonstration of compliance.
What about penalties?
Failure to comply with these transparency obligations may result in fines of:
• up to €15 million or 3% of the company’s total annual worldwide turnover, whichever is higher, with proportional application to SMEs and small mid-cap companies;
• up to €750,000 in the case of EU institutions, bodies, offices and agencies.
The application of these rules represents another step towards building a European framework in which technology and innovation can develop while ensuring due respect for users’ fundamental rights.
