On November 14, 2025, a draft law amending the Federal Law “On Information, Information Technologies and Information Protection” was submitted to the State Duma for consideration, along with a draft law amending the Code of Administrative Offences.
Proposed Amendments:
- Definition of Synthetic Video Material. Synthetic video material is proposed to be defined as visual material that is fully or partially created, altered, or processed using artificial intelligence technologies, including generative neural networks and machine learning algorithms, resulting in a visual image that mimics real-world objects, faces, events, or actions.
- Labeling of AI-Generated Video Content. The draft law stipulates the obligation for owners of social networks, video hosting services, and other internet resources to label video content with the captions “Created with the use of artificial intelligence” or “AI-Generated,” as well as to display a machine-readable label in the video file’s metadata containing information on the use of AI technologies, the label creation date, and the resource owner’s identifier. The label must be visible throughout the entire duration of the video material.
- Definition of a “Video Hosting Owner”. A video hosting owner is defined as the owner of a website and/or a website page on the Internet, and/or an information system, and/or a computer program intended for or used for the placement, storage, reproduction, and distribution of video materials by Internet users. In other words, the entities subject to liability will be platforms with user-generated content functionalities (e.g., YouTube and similar hosting services), social networks, and other services where video content can be posted.
- Administrative Liability. The text of the draft law explicitly states the necessity of introducing administrative liability for violations of the AI-content labeling rules. The proposed fine amounts are:
- for individuals: from 10,000 to 50,000 RUB;
- for officials: from 100,000 to 200,000 RUB;
- for legal entities: from 200,000 to 500,000 RUB.
It is worth noting that the proposed amendments align with international regulatory trends.
- European Union. Requirements for AI-content labeling are in effect at the EU level under the AI Act (adopted in 2024, coming into force in 2026). However, the prescriptions of the European regulator are somewhat broader and provide for mandatory labeling not only for video content but also for generated audio, text, sound, and other content. Mandatory labeling applies to any AI-content made available to EU citizens, even if it was created outside the EU and regardless of the platform’s country of origin.
- China. As of September 1, the law “On Measures for the Management of Generative AI Services” has been in effect, also establishing AI-content labeling requirements. It is important to note that the law sets a higher standard of conduct for online platforms: if a platform discovers content that was potentially created via AI generation, it must notify the user and, if necessary, add the label itself.
- United States. The COPIED Act (Content Origin Protection and Integrity from Edited and Deepfaked Media Act) is under consideration in the US Senate – a draft law on the identification and regulation of content created with the help of AI.
Some states have already introduced their own AI-labeling regulations. For example, in California, law SB 243 was passed in 2025, which obligates chatbot operators to ensure transparency about their artificial nature and to implement safety measures, especially for underage users.
The introduction of labeling for generated video content is of particular importance for the advertising market, which is the most active user of AI for producing commercials. Currently, it can be assumed that if the law is adopted in its current version, video hosting owners, being the liable entities, will most likely introduce internal requirements for video content labeling and update moderation rules to identify AI-generated materials. This means that video content creators publishing videos on such platforms will be obligated to comply with the new labeling standards.
Separately, it is important to note the necessity of conducting an audit of already published AI-video content for preventive purposes.
Some major companies are already independently introducing rules for the voluntary labeling of AI-content. For instance, Yandex, in collaboration with the Institute for System Programming of the Russian Academy of Sciences, is working on developing a system for labeling AI-generated content.