The EU will support industry efforts to slow the development of more disruptive AI technologies and will collaborate with partners such as Canada and the UK on AI safety, the Commission chief said on Wednesday.
In her annual State of the Union address to the European Parliament, European Commission President Ursula von der Leyen hailed AI's potential to benefit humanity, provided its risks and opportunities are properly balanced.
She singled out the growing risks posed by so-called frontier AI models, those pushing the boundaries of the technology.
"As the frontier models become more capable, these risks have come sharper into focus. Models being developed will allow hacking on a level we never thought possible. And they will soon be in the hands of adversaries who see the world very differently to us," she said.
The Commission President referred to recent incidents in which AI agents escaped their testing environments and independently hacked other systems, as occurred in the OpenAI-Hugging Face breach in July.
"CEOs of the most advanced companies tell us that it is time to slow down on the self-recursive models. To pace the frontier. If the people developing the technology are clear, then we should be too," von der Leyen continued.
The EU already has one of the world's most comprehensive laws on artificial intelligence, the AI Act, which gives the Commission oversight of risk-mitigation measures for the most advanced models.
"We will step up together with our like-minded partners like Canada, the UK and others. We want to team up on model evaluation, verification, early-warning, AI security and much more," she added.
The Commission chief also announced she will convene a discussion with leading AI labs on how public authorities can support industry efforts to pace the frontier of this disruptive technology.
