The European Union has made little progress in persuading Anthropic PBC to allow testing of banks and companies for digital vulnerabilities revealed by the Mythos AI model, according to Spain’s economy minister Carlos Cuerpo [1, 2]. Mythos is an AI tool designed to identify unknown IT system flaws and has raised concern across Europe as Anthropic restricts access to it [2].
Carlos Cuerpo described the stalled negotiations as “one of the most important elements not only in the medium and long term, but also in the short term,” emphasizing financial stability as a global issue that transcends national borders. He said disruptive models like Mythos could affect banks not only in Europe and the UK, but also Canada and the United States due to the interconnected nature of the global financial system [2]. “This is not something that affects only one country,” he stressed.
The issue was due to be raised again at the EU finance chiefs meeting in Nicosia on May 22, 2026, where Cuerpo said talks with Anthropic had seen only limited progress so far [1, 2]. European authorities and international bodies such as the G20 and the Financial Stability Board may need to get involved given the global impact of the vulnerabilities Mythos could expose [2].
Anthropic has reportedly begun working with U.S. companies to identify potential risks posed by the Mythos AI tool [2]. However, European officials remain concerned due to restricted access to the model and the potentially wide-reaching financial risks it could reveal.
With little advancement in EU-Anthropic discussions before the May 22 meeting, attention will focus on whether the finance chiefs can push further action to address the emerging digital threats uncovered by the Mythos AI model.
