Orange Cyberdefense joins Mistral AI design partner program

A new partnership for security

Orange Cyberdefense has officially joined the design partner program for Mistral Large 4. This collaboration marks a significant move for the French security firm as it seeks to enhance its defensive capabilities. By partnering with Mistral AI, the company aims to integrate advanced artificial intelligence into its core security offerings.

The firm will use the new model to help test client systems for weaknesses and investigate active breaches. The goal is to provide faster responses to cyberattacks by leveraging the processing power of the AI. This initiative represents a practical application of large language models in the high-stakes world of cybersecurity.

As reported by The Next Web, Orange Cyberdefense is the first European managed security services provider to participate in this specific program. This distinction highlights the importance of the partnership for the European tech landscape. The company plans to introduce the model across its various services in a gradual, phased approach.

How the AI will be deployed

The first practical uses for Mistral Large 4 will involve technical audits and penetration testing. In these scenarios, security specialists intentionally try to break into a client's systems to uncover vulnerabilities. The AI will help these experts identify the specific paths an attacker might take to gain unauthorized access.

Beyond proactive testing, the model will also be used to investigate incidents after a breach has occurred. Security teams can use the AI to analyze what happened during an attack and how the intruders moved through the network. This capability is designed to help staff understand the scope of a breach and respond more quickly to mitigate damage.

The model itself is a large-scale system with one trillion parameters. Mistral launched this model in early October and has made its weights available for download. This allows developers to run and adapt the model themselves, which is a key feature for companies that require custom configurations for their specific security needs.

Prioritizing data sovereignty

One of the most critical aspects of this partnership is where the data actually lives. Orange Cyberdefense will run its security models on computers located in Mistral's European data centers. This setup is specifically designed to help clients retain control over their sensitive information during security work.

Data control is a major consideration for organizations when they decide to use AI for sensitive tasks. By keeping the data within European borders and using a model where the weights are accessible, the firms can offer more transparency. Mistral has positioned deployment flexibility as a primary argument for using open models over closed systems.

Anis Trabelsi, head of innovation and AI at Orange Cyberdefense, stated that the alliance strengthens the foundation of cybersecurity augmented by trustworthy AI. This focus on trust is essential for public-sector and large business clients who cannot risk their private data being used to train external models. The arrangement ensures that the AI serves as a tool for the client rather than a risk to them.

Expanding existing AI tools

This partnership does not represent the first time Orange Cyberdefense has utilized artificial intelligence. Since April 2025, the firm has already been using a tool from Qevlar AI within its security operations centers. These centers are the hubs where staff monitor client systems around the clock for potential threats.

The addition of Mistral Large 4 builds upon this existing technological foundation. While the Qevlar AI tool serves a specific role in monitoring, the Mistral model provides a broader suite of capabilities for testing and analysis. The company is essentially layering different AI technologies to create a more robust security posture for its customers.

The financial details of the agreement between Orange Cyberdefense and Mistral AI have not been disclosed. However, the scope of the rollout is clearly defined. The company will initially target its business and public-sector clients, who often have the strictest requirements for both security and data privacy.

The future of open models

The move by Orange Cyberdefense highlights a growing trend toward using open models for enterprise security. Because Mistral has made the weights for Large 4 available, developers can modify the model to suit specific security protocols. This level of customization is often impossible with closed AI systems where the underlying architecture is hidden.

By testing a cybersecurity-focused version of the model with a small group of partners, Mistral is creating a feedback loop for real-world applications. These partners are restricted to using the model only on systems they own or have permission to assess. This ensures that the development of the AI remains safe and controlled during the testing phase.

As more managed security providers adopt these models, we can expect to see a shift in how penetration testing is conducted. The ability to simulate complex attacker behaviors at scale could become a standard offering. For the reader, this means that the security of their digital infrastructure may soon be defended by models that can think and react with greater speed.