EU to discuss access to Mythos AI model with Anthropic

European Commission pushes to ensure access to Mythos Preview for European companies
US authorities recently blocked wider distribution citing security risks
Mythos Preview can reveal decades-old vulnerabilities and generate exploits quickly.

The European Commission (EC) is trying to secure access to Anthropic’s AI model, Mythos Preview, for European companies and is sending emissaries across the Atlantic to try to lobby for such an outcome.

Citing unnamed sources familiar with the matter, Bloomberg said the EC would send a few officials to San Francisco this week to meet with Anthropic PBC representatives to learn more about the tool and try to make it available to the bloc.

The publication said the EU has been pushing for access since Anthropic first revealed Project Glasswing, seeking to test its networks, as well as those of EU banks, critical infrastructure companies and technology companies. Apparently, Anthropic also wants to increase the number of organizations that are part of the trial phase, but it is being held back by the US government.

The Glasswing project is important

“White House officials recently rejected Anthropic’s plans to distribute Mythos to several dozen additional companies and organizations, citing security concerns,” the publication writes. At the same time, French ministers “demanded” access for EU banks and businesses, it was also said, and in early May Eurogroup President Kyriakos Pierrakakis highlighted the importance of Mythe:

“I don’t think we have the luxury of not trying to establish channels of communication with the United States.” » said Pierrakakis. “The challenge here is that technologies like AI require international governance frameworks at a time when multilateralism is being challenged. »

Anthropic’s announcement of the Mythos Preview model shook the entire cybersecurity industry. Apparently, the tool can easily reveal decades-old vulnerabilities in fully patched systems and programs and use them to create working exploits. AI was deemed too dangerous to share with the public and was handed over to a handful of key organizations (banks, critical infrastructure companies, etc.) so they could secure their products before bad actors got their hands on the tool.

Via Bloomberg