The European Union is well prepared to address potential risks from increasingly powerful artificial intelligence systems, European Commission Executive Vice-President Henna Virkkunen has said, pointing to the bloc’s AI Act as a framework for overseeing the technology.
Speaking on Friday, October 9, Virkkunen said the EU’s rules cover the lifecycle of AI models, including safety assessments, expert involvement and continued monitoring as the technology develops.
Her comments come amid growing international concern about whether advanced AI systems could behave in unexpected ways or operate beyond effective human control. Recent incidents involving AI companies, including OpenAI and Anthropic, have added to debate over how developers test and manage powerful systems.
The EU AI Act uses a risk-based approach, imposing different obligations depending on how an AI system is used and the level of risk it may pose. The legislation is intended to strengthen safety and accountability while allowing beneficial applications of the technology to continue developing.
Virkkunen rejected criticism from some technology companies that the rules are already outdated. She said EU lawmakers had anticipated further advances in AI and included provisions for ongoing assessments and input from external experts.
The European Commission has also increased its scrutiny of AI developers. According to Reuters, it sent requests for information to more than 30 AI companies in late August, seeking details about their safety and security measures as well as their transparency and copyright practices.
The companies approached were not publicly named. Virkkunen said the Commission was assessing their responses and had also requested information from Chinese AI developers.
Those requests could lead to formal investigations and, where a breach is established under applicable rules, financial penalties. Reuters reported that fines could reach up to 7% of a company’s global annual turnover.
The debate over AI safety is unfolding alongside competition between the United States, China and Europe to develop advanced systems. Virkkunen acknowledged that the most capable models currently available largely come from US and Chinese companies, adding that their capabilities bring risks regulators must understand.
She also addressed discussion of a possible digital levy that would mainly affect large US technology companies. Virkkunen said decisions on new sources of funding for the EU’s 2028–2034 budget rest with member-state governments, which are considering different options.
The EU’s approach will face continuing tests as AI capabilities evolve. Although the bloc believes its existing framework provides tools to address emerging risks, its effectiveness will depend on implementation, enforcement and the ability of regulators to respond to new developments.