Brussels Tightening Oversight of Artificial Intelligence

The European Union is moving into a new phase of regulating artificial intelligence technologies. On August 2, the main provisions of the EU Artificial Intelligence Act came into effect, including requirements regarding the transparency of AI systems and the content they generate. Developers are now required to clearly inform users about the use of artificial intelligence when providing services or creating content.

The tighter regulations are driven first of all by the rapid spread of deepfakes, disinformation, and other manipulative content. Failure to comply with the requirements is punishable by a fine of up to EUR 15 million or 3% of the company’s global annual turnover. The new rules apply to AI systems entering the EU market for the first time, while a four-month transition period has been established for existing services. Content created by users exclusively for personal purposes is not subject to these requirements.

Chatbots and other interactive AI systems must inform users that they are not interacting with a human. Mandatory labeling also applies to deepfakes: such content must include both a visible label and a machine-readable tag for automatic detection, such as digital watermarks. Texts on topics of public importance created by AI without human editorial oversight are also subject to labeling.

Oversight of compliance has been entrusted to the European Commission’s Office for Artificial Intelligence (37 staff members), which, in collaboration with national authorities in EU member states, will assess the safety of the most powerful models, require companies to take additional measures, gain access to technical documentation, and conduct inspections. In case of violations, the Office will be able to impose fines and require that models be restricted or withdrawn from the market.

Leading developers, including OpenAI, Google, and Meta, have expressed their willingness to cooperate with the European Commission and have signed on to the Code of Good Practice developed to implement the provisions of the Act. The document has been endorsed by more than 180 companies.

The adoption of uniform, mandatory rules cements the European Union’s status as one of the world’s leading centers for the legal regulation of artificial intelligence. The introduction of these regulations marks a shift from reacting to isolated instances of disinformation to systemic countering digital and hybrid threats, as the mandatory labeling of AI-generated content will have not only technological but also security implications for protecting electoral processes and public trust in government institutions.