EU tech chief Virkkunen says AI rules can tackle rogue systems

She said regulators were providing companies with guidance on evaluating their models and determining the time required for assessments. Recommendations from a scientific panel comprising 60 experts from universities and institutions would also inform the process.

Virkkunen dismissed criticism from some technology companies that the EU's AI rules were already outdated, saying lawmakers had anticipated emerging developments by requiring risk assessments, external expert involvement and continued monitoring of AI models. In late August, the European Commission sent requests for information to more than 30 AI companies, seeking details about their safety and security measures.

It subsequently sought further information about their transparency and copyright obligations. The requests could lead to investigations and fines of up to 7% of a company's global annual turnover. The Commission has not identified the companies contacted, and Virkkunen said she was assessing their responses.

Asked about Chinese AI startups, she confirmed that the Commission had also requested information from them, adding that the most capable AI models currently available came from the United States and China. Virkkunen acknowledged that increasingly capable models were creating new risks that regulators needed to monitor closely.

On the possibility of introducing a digital levy that would primarily affect major US technology companies, she said the decision rested with EU governments, which are discussing new sources of revenue for the bloc's 2028-2034 budget.