Перейти к содержанию

AI-generated content can no longer be hidden: The Russian Union of Journalists demands every AI-generated material be labeled

The document mandates disclosure of AI involvement and assigns responsibility to humans

The Russian Union of Journalists (SJR) has developed recommendations for newsrooms on the use of artificial intelligence, TASS reports. The list covers five key areas: content labeling, journalist responsibility, synthetic visual content, internal editorial processes, and interaction with platforms.

The main requirement is a mandatory disclaimer at the beginning or end of the material, disclosing the fact of AI use. The labeling must be informative: it should indicate the specific role performed by the algorithm (draft generation, translation, editing) and whether a human participated in the verification. For topics related to life, health, finance, and safety, the requirements for verification and labeling are made as strict as possible.

It is separately emphasized that AI remains only an auxiliary tool. Personal responsibility for facts, quotes, and data is proposed to be assigned to the journalist and the editorial office, not to the algorithm. This approach should protect the audience from AI errors and maintain trust in the media.

The recommendations are essentially advisory in nature, but they effectively set a new standard for an industry that is increasingly using automation in material preparation.

Context: How AI content labeling is regulated in Russia

The topic of labeling AI-generated content is actively discussed at the legislative level. As early as April 2025, the State Duma considered a bill on AI regulation, where mandatory labeling was called the first and most obvious step — considering risks to health, safety, and copyright. Later, the head of the information policy committee, Sergey Boyarsky, directly proposed obliging media and platforms to indicate whether the text was written by a human or an algorithm.

In parallel, a broader discussion about labeling approaches developed. In March 2026, the first deputy chairman of the same committee, Anton Gorelkin, expressed an unconventional idea: since AI content already constitutes the majority of social media publications, it would be more reasonable to label not it, but materials created without AI — by analogy with farm products.

Read more on the topic: