European Commission President Ursula von der Leyen said that the EU wanted “to deepen” its cooperation with the United States in the field of artificial intelligence (AI), following the Trump administration’s decision to restrict access to advanced AI models from Anthropic, Mythos and Fable, to people who are not US nationals (see EUROPE 13888/14).
“We want to invest together and ensure that our citizens and businesses benefit from the best possible protection against malicious actors”, she added during the working lunch devoted to AI, organised as part of the G7, on Thursday 18 June, according to the transcript of her remarks released by her team. She insisted on the “mutual interest” of both parties in “safely using the best artificial intelligence models”.
As early as 16 June, Ms von der Leyen had already described cooperation on AI at G7 level as “essential” in order to “exchange information and ensure cooperation between the independent entities responsible for evaluating models before they are placed on the market”, while insisting that the EU wanted to build “its own future in AI” with these “trusted partners”, thanks to the tech sovereignty package (see EUROPE 13880/1).
According to sources consulted by Agence Europe, Anthropic’s chief executive argued during the G7 discussions in favour of the EU being granted access to the company’s models. Representatives of ENISA and of Anthropic are meeting today in San Francisco to discuss the company’s proposal to grant access to its Mythos model (see EUROPE 13879/14). (Original version in French by Ana Pisonero Hernández)