In a nutshell: After weeks of exclusion, the EU now gains access to Anthropic’s Mythos hacking AI model, while the Commission works on a formal action plan for such systems.
Anthropic has invited the European Commission and its cybersecurity agency ENISA to access its AI model Mythos, which automatically detects and exploits security vulnerabilities. EU authorities had previously been denied access to this technology for weeks and urgently called for action.
Anthropic has formally invited the European Commission to access its AI model Mythos through its cybersecurity agency ENISA. This follows a meeting between the Commission and Anthropic in San Francisco, after which the EU must now establish security mechanisms for access. An ENISA spokesperson confirmed that the agency does not yet have active access, but is working on implementation.
Anthropic introduced Mythos in early April and warned that the model outperforms the vast majority of people in detecting and exploiting cybersecurity vulnerabilities. This raised questions about the security of critical infrastructure should the model fall into the hands of cyber attackers. European politicians and government officials subsequently spent weeks urgently demanding access to the technology or calling on the EU to develop its own version.
EU spokesman Thomas Regnier stated that the Commission had “conducted several productive meetings with Anthropic” and “welcomed” the new developments regarding potential future access. He emphasized: “Mythos is not unique; a new wave of powerful models is coming to market.” The Commission is working on a formal action plan to respond to such AI hacking tools and intends to publish it before the summer recess.
Source: www.politico.eu · Published June 1, 2026
Lumi AI News — AI-assisted curation pursuant to Article 50 EU AI Act. Paraphrase and classification by Lumi News Pipeline v1.2.8.