By Edith Hancock
The European Union proposed a bundle of measures to address the risks of advanced artificial intelligence tools and manage access to frontier models used in cybersecurity in the bloc as Europe trails the U.S. in harnessing the fast-evolving technology.
"AI is transforming the meaning of cybersecurity. And we must keep pace," Henna Virkkunen, the European Commission's top tech official, said on Tuesday.
Under the plan, the EU's executive arm wants to ramp up evaluations of frontier AI models, develop a testing platform for AI with advanced cyber capabilities and set out a blueprint for giving European organizations access to advanced AI tools for cybersecurity uses.
The so-called action plan is aimed at strengthening both governments' and companies' ability to address risks that AI models pose and their own advanced AI cyber capabilities.
European institutions are trying to use cutting-edge technologies from companies such as Anthropic for security research. The commission is currently in talks to gain access to Anthropic's Mythos AI model.
Write to Edith Hancock at edith.hancock@wsj.com