According to Bloomberg reports on May 29, European Commission officials plan to travel to San Francisco to meet with representatives of Anthropic PBC to learn more about the Mythos model details and request access for the EU. Insiders said the officials are scheduled to visit San Francisco on Thursday, and all insiders asked to remain anonymous because the plan is still being formulated.
EU seeks access to Mythos: confirmed background
Since Anthropic revealed that the Mythos model has outstanding capabilities in identifying network vulnerabilities, and that it could pose a major cybersecurity risk, the EU has been trying to obtain access. The insiders cited by Bloomberg said that such access may require approval from the U.S. government to be released.
Anthropic has also confirmed plans to widely release a new model in the coming weeks, with network security capabilities comparable to Mythos.
Opening of Anthropic’s Milan office and expansion in Europe
According to Reuters on May 28, Chris Ciauri announced that Anthropic has opened its sixth office in Europe in Milan. Initially, it will focus on sales, marketing, and providing pre-sales and post-sales technical support for customers such as Generali, Unipol, Pirelli, Bending Spoons, and Satispay.
Anthropic plans to double the number of international employees. The London office is set to expand to about 800 employees (publicly announced in London a few weeks ago). Anthropic’s existing offices in Europe include Dublin, Zurich, Munich, Paris, and London (London currently has about 200 employees).
Confirmed external relationship background for Anthropic
Anthropic co-founder Christopher Olah is the only technology-company representative invited to attend this week’s release event of Pope Francis’ first encyclical (involving an AI challenges theme). In addition, Anthropic has previously clashed with the Trump administration over restrictions on model use, especially limiting its model from being used for military purposes such as autonomous weapon targeting and domestic surveillance.
FAQ
Why is the EU particularly concerned about Anthropic’s Mythos model?
Since Anthropic revealed that Mythos has outstanding capabilities in identifying network vulnerabilities and that it could pose a major cybersecurity risk, the EU has been trying to obtain access. By contrast, OpenAI has opened GPT-5.5-Cyber for the EU, while Anthropic has not yet provided Mythos.
Does accessing Mythos require U.S. government approval, and where do things stand?
According to insiders cited by Bloomberg, access to the Mythos model may require approval from the U.S. government to be released. At present, the European Commission is planning to send officials to visit San Francisco to advance negotiations, but because the plan is still being formulated, specific details have not been disclosed.
What business areas does Anthropic’s Milan office focus on?
Per Chris Ciauri’s statement, initially the Milan office will focus on sales, marketing, and providing pre-sales and post-sales technical support for Italian customers such as Generali, Unipol, Pirelli, Bending Spoons, and Satispay, while arranging interactions with policy makers locally.
