Orange Cyberdefense struck a partnership with French player Mistral AI to develop services using a specialised version of its latest large language model, as European companies look to strengthen defences against automated digital threats.

The deal gives Orange Cyberdefense early access to a cybersecurity-optimised version of Mistral Large 4. The companies aim to apply the technology across multiple stages of defence operations.

Initial work will focus on penetration testing of information systems and technical audits, using AI to explore a wider range of potential attack scenarios and identify weaknesses in customers’ systems. The technology will also be used to speed post-incident investigations by analysing security traces and reconstructing attacks.

Both companies stressed use cases would operate under human oversight, with access management, traceability and security testing among the safeguards. Critical actions will remain the responsibility of experts.

The Orange unit also spotlighted sovereignty, stating it would use European computing capacity to run the cybersecurity models.

Mistral AI co-founder and CTO Timothee Lacroix noted the companies want to provide European organisations with a “sovereign, trusted AI solution to defend themselves”.

The partnership comes as cybersecurity providers contend with hackers increasingly using AI to automate attacks.

“Volume, speed, trust: these are the three challenges facing cybersecurity in the age of AI,” noted Orange Cyberdefense CEO Hugues Foulon.

“As threats become ever faster, industrialised and now automated, defence must operate at a different scale,” he added, arguing that combining AI capabilities with its threat intelligence would help protect customers “more broadly, more quickly and with trusted European technology”.

The post Orange unit inks cybersecurity pact with Mistral AI appeared first on Mobile World Live.