Nova pravila EU: vsebine, ki jih ustvari umetna inteligenca, morajo biti označene

The European Union has begun implementing new rules requiring content generated by artificial intelligence to be clearly marked, starting August 2. The regulations, part of the AI Act, aim to enhance transparency and reduce risks associated with AI-generated material. Under these guidelines, interactive systems such as chatbots must inform users they are communicating with AI rather than humans. Additionally, images, videos, and audio content created or modified by AI, including deepfakes, must be labeled. These labels will include machine-readable tags to facilitate detection and ensure clarity for users. The implementation follows the publication of a list containing over 180 organizations that have signed the Code of Good Practice for Transparency of AI-Generated Content. This initiative underscores the EU's commitment to fostering responsible AI development while safeguarding citizens against potential manipulation and misinformation. The AI Act introduces stricter requirements for providers of general-purpose AI models, which can be used across various tools and services. These entities must document specific information and submit it to relevant authorities or other providers. They are also required to establish copyright protection policies and disclose detailed summaries of the content used to train their models. Particularly harmful AI systems are now banned under the new regulations. These include systems that manipulate individuals, exploit vulnerabilities, or unfairly assess people in ways that threaten their rights. The AI Office will enforce rules for systems offered by the same provider as the foundational general-purpose model and those integrated into large internet platforms and search engines. National authorities will handle other systems, while the European Data Protection Supervisor will oversee systems used by EU institutions, bodies, and agencies. A Scientific Committee composed of 60 independent experts will assist in the enforcement process. Professor Alessandro Abate from the University of Oxford has been appointed as the AI Office’s chief scientific advisor. A reporting tool is available for citizens and businesses to file complaints regarding alleged violations of the AI Act. Individuals working with AI system or model providers can securely report possible infringements through a dedicated whistleblower channel. A special channel is also available for providers developing their own systems using general-purpose models. The AI Office has stated that all information received through these channels will be handled confidentially. Implementation of rules for high-risk AI systems has been postponed until December 2, 2027, while regulations for such systems integrated into regulated products will apply from August 2, 2028. New bans on systems generating sexually explicit content or materials involving child sexual abuse without consent will take effect on December 2, 2026. These measures reflect the EU's ongoing efforts to balance innovation with ethical considerations and public safety.