They will be divided into three classes. The first will include publicly available data, the second – anonymized confidential, including those related to banking, medical, tax secrets, and the third – data containing state secrets.

Deliberately created state operators will provide access to such sets based on a motivated request. In this case, the cost of creating one set will be at least 1 million rubles. The composition and sources of data for datasets will be determined by the state operators themselves, who will also be obliged to depersonalize information and provide access to it, depending on the security class. According to the draft roadmap of the Ministry of Digital Development, access to different classes of data will be opened in 2022–2024.

In August, it was reported that a mode of access to datasets for AI developers appeared in the R&D Infrastructure catalog.

Original (in Russian)