regulation of artificial intelligence in the EU
Meta has officially refused to sign the European Union Code of Conduct for Regulating Artificial Intelligence, expressing concern about the potential negative impact of the document on innovation and technology development in the region. The announcement comes a few weeks before new pan-European regulations on suppliers of general-purpose AI models come into force.
This is a code developed by the European Commission as part of the implementation of the Law on Artificial Intelligence. The document provides for voluntary but recommended measures aimed at increasing transparency, copyright compliance, and ensuring ethical standards in the development and use of AI systems. Among the obligations are regular updating of documentation on AI tools, a ban on the use of unlicensed content for training models, as well as mandatory compliance with the requirements of copyright holders to exclude their works from training datasets.
Meta management considered the proposed measures to be excessively strict and legally vague. Representatives of the company believe that the document goes far beyond the approved legislation and can significantly slow down the development of advanced AI models in Europe. Of particular concern were the risks for technology companies developing business models based on generative technologies.
AI regulation in the EU is based on a risk assessment model. It prohibits a number of use cases that are classified as "unacceptable risk" — for example, behavioral manipulation or the creation of social rating systems. "High-risk" areas are also identified — biometrics, facial recognition, technologies in education and hiring, where developers are required to register systems and ensure risk and quality management.
The new regulations have sparked intense discussion in the technology community. Major companies-from American corporations to European startups-have previously called on regulators to delay the implementation of the law's provisions. However, the European Commission has made it clear that the deadlines will be met. As early as August 2, obligations will come into force for developers of models with systemic risks, which include, among others, OpenAI, Google, Anthropic and Meta.
In parallel, on July 18, the European Union issued separate recommendations on how to prepare for compliance with the new regulation. Their mission is to help technology companies adapt to requirements, minimize legal costs, and maintain the industry's growth rate.
The conflict over the Code of Conduct highlights the growing tension between the EU's ambitions to create a regulated digital environment and the interests of multinational technology corporations. The question of the balance between protecting citizens ' rights and promoting technological progress is becoming increasingly relevant, especially in the context of the rapidly developing artificial intelligence sector.
25 июля 2025 года с российского космодрома Восточный запланирован запуск ракеты-носителя «Союз-2.1б», которая доставит на околоземную орбиту телекоммуникационный спутник Nahid-2,…
Американская компания Starlink запустила спутниковый сервис T-Satellite — телекоммуникационное решение, ориентированное на обеспечение мобильной связи в зонах, где отсутствует покрытие…
Американская корпорация Microsoft сообщила о достижении нового уровня стабильности в своей операционной системе: Windows 11 24H2 признана самой надёжной версией…
Американская компания SpaceX рассматривает возможность использования ракетной системы Starship не только для космических миссий, но и в качестве средства сверхбыстрого…
Компания xAI, основанная Илоном Маском, представила амбициозный проект в области высокопроизводительных вычислений — суперкластер Colossus 2, который станет первым в…
Американский режиссёр Джеймс Кэмерон завершает работу над третьим фильмом знаменитой франшизы «Аватар». Картина под названием «Аватар 3: Огонь и пепел»…