🎓 Free Web Scraping for AI Training Sanctioned by New Legislation

The amendments, key provisions of which come into force on September 1, 2026, exempt Big Tech firms from royalty payments to copyright holders, provided the data was obtained “lawfully” and made publicly available. Industry analysts estimate this will save developers of foundational models—such as YandexGPT, GigaChat, and T.Pro 2.0—hundreds of millions of rubles while accelerating AI integration across sensitive economic sectors.

Conversely, representatives of creative industries, publishers, and record labels have fiercely criticized the measure, deeming it a devaluation of human authorship. Citing research from the Institute for Statistical Studies and Economics of Knowledge (ISSEK HSE), publisher Eksmo warned that cumulative income losses across 11 key creative professions from unchecked generative AI adoption could exceed 60 billion rubles by 2030, with broader downstream economic damages reaching 1 trillion rubles. The National Federation of the Music Industry (NFMI) projects annual direct losses for songwriters at 1.5 billion to 4 billion rubles, anticipating a 30% to 35% decline in original content valuation.

To mitigate these emerging risks, digital book services and music distributors plan to implement technical protection measures—ranging from anti-scraping metadata tags (opt-out mechanisms) to mandatory watermarking of AI-generated media across streaming platforms.

Legal experts note that the statute functions as a framework agreement, leaving the creative sector a window of opportunity to lobby for compromise via secondary legislation before March 2027 by pushing for mandatory licensing regimes and commercial data usage royalties.