AI-Generated Anchors and Texts: Who Will Control AI Use in Serbia's Media?

Serbia is lagging behind the European Union in regulating artificial intelligence, particularly its use in media. While the EU mandates clear labeling of AI-generated content, Serbia's law is still in preparation, leaving a regulatory vacuum. This "digital Wild West," as described by one expert, allows for the unchecked appearance of AI-generated content, including "deepfakes," in media and on social networks.

“

If a media house uses deepfake content to create a certain report or other media content, it must clearly inform the audience that they are watching or listening to content generated or modified using artificial intelligence.

Recent examples highlight the issue: Happy TV has featured AI-generated anchors, and Požarevac's SAT TV used AI presenters for a project funded by the Ministry of Information. These instances raise concerns about transparency and the potential for misinformation. Experts note that while existing Serbian media laws prohibit misleading citizens, the lack of specific AI regulations creates loopholes.

“

The state foresees a special agency under the Government's supervision that will deal with AI use issues. But I emphasize that the Law on Public Information and Media already has rules concerning truthful, timely, and responsible reporting, which would apply to media, or unfortunately, tabloid media and media outlets, that use such tools for propaganda purposes.

The Independent Association of Journalists of Serbia (NUNS) is skeptical that new laws will deter tabloids already using AI for propaganda. Rade Đurić, a legal expert at NUNS, stated that while the Journalists' Code of Serbia recognizes AI, its enforcement is inconsistent. He expressed doubt that new rules will be applied to those determined to break them, questioning the state's effectiveness in imposing sanctions.

“

The Code of Journalists of Serbia already recognizes artificial intelligence, however, looking at how the Code is applied and who simply does not want to apply it, we think they will not apply the rules related to artificial intelligence either. On the other hand, the way the state reacts, or how it prepares an adequate response in the form of processing and certain sanctions, we are not sure that it will affect those who are determined to break these rules.