EU Commission Approves AI Content Labeling Symbols

The European Commission has adopted symbols intended to clearly mark artificial intelligence-generated content, ensuring users can easily identify it. These labels must be noticeable and not obscured by other elements, remaining visible even when content is shared or downloaded.

In Lithuania, the Radio and Television Commission (LRTK) views enhanced AI content labeling as a crucial step in combating disinformation. "AI-created images or voices can already appear so convincing that it becomes increasingly difficult for people to distinguish fakes themselves," stated LRTK chairman Mantas Martišius. He emphasized that the responsibility cannot solely rest on users to identify AI-generated material, making clear labeling essential.

“

AI-created images or voices can already appear so convincing that it becomes increasingly difficult for people to distinguish fakes themselves. Therefore, we cannot shift all responsibility to the user and expect them to recognize what is real and what is created by artificial intelligence. Labeling must convey that information for them.

However, the LRTK acknowledges practical challenges in enforcing these requirements, particularly when AI tools and platforms operate outside the European Union. "The EU does not really create its own AI tools. We can stipulate, demand implementation, but if these AI tools are outside the EU, and the vast majority are, the question of how control will be exercised arises," Martišius noted.

Approximately 190 organizations, including tech giants like Google, Meta, Microsoft, and OpenAI, have joined the voluntary code of practice for AI transparency. Lithuanian media outlets such as Delfi, Lrytas, Bernardinai.lt, and 15min have also signed on. Nevertheless, the transparency requirements mandated by the EU AI Act apply independently of code adherence, and the use of the EU-created symbols is voluntary. The AI Act, which entered into force in August 2024, aims to mitigate risks of deception and manipulation, with a particular focus on deepfakes and other AI-generated content that could be used for misinformation or fraud.

“

The EU does not really create its own AI tools. We can stipulate, demand implementation, but if these AI tools are outside the EU, and the vast majority are, the question of how control will be exercised arises.