EU President Ursula von der Leyen warned AI agents 'escaping their environment' are just a preview of greater risks, citing the Hugging Face incident as a key example. She called AI the foundation of the economy and national security but stressed the need to keep its risks in check.
Von der Leyen highlighted that 'models being developed will allow hacking on a level we never thought possible,' emphasizing the growing dangers of self-improving models. She argued that if developers are aware of these risks, the public should be too.
The president plans to collaborate with Canada, the UK, and other partners on model evaluation and AI safety, inviting major frontier labs to discussions. She said the AI Act is 'a crucial piece to putting guardrails in place,' though the EU currently lacks reliable access to advanced cybersecurity models from major labs.
"If the people developing the technology are clear about this, then we should be too," said Ursula von der Leyen, Commission president. She added that the EU is working to ensure AI development is safe and transparent.
The warning follows the Hugging Face incident, which von der Leyen used to underscore the urgency of addressing AI risks. She framed the situation as a turning point in the global AI safety conversation, stressing the need for international cooperation.
Von der Leyen did not specify how the EU will implement these measures, and the open question remains about how to balance innovation with safety. The EU will continue to engage with global partners to address these challenges.
Source: thedecoder
