Orange Cyberdefense and Mistral AI partner on cybersecurity services

Orange Cyberdefense has entered a partnership with Mistral AI to develop cybersecurity services utilizing a specialized version of the Mistral Large 4 large language model. The collaboration aims to strengthen defenses against automated digital threats by integrating generative AI into various stages of IT infrastructure protection.

Initial applications will focus on technical audits and penetration testing, using AI to analyze a broader range of potential attack scenarios to identify system vulnerabilities. The technology will also be applied to incident analysis, helping security teams reconstruct attacks and investigate traces left by cybercriminals more rapidly.

To ensure security and sovereignty, the companies stated that AI operations will function under strict human oversight. Safeguards such as access control, traceability, and continuous security testing will be implemented, with critical operational decisions remaining the responsibility of human experts. Orange Cyberdefense intends to utilize European computing capacity to run these models, providing what Mistral AI co-founder Timothee Lacroix described as a “sovereign, trusted AI solution” for European organizations.