CyberSecurity, Offensive Security Engineer

CyberSecurity, Offensive Security Engineer Overview

This role sits within the engineering and infrastructure organization and focuses on offensive security for Mistral AI’s agentic products and supporting systems. The work is centered on protecting advanced AI experiences by thinking like an attacker, finding weaknesses early, and helping the company build more resilient products, infrastructure, and defenses. The position is especially relevant for someone who enjoys operating at the intersection of offensive security, AI safety, and product development in a fast-moving environment.

What the role involves

• Searching for vulnerabilities across the interactions between agent-based applications, cloud infrastructure, and foundational models, with an emphasis on realistic attack paths that could have serious impact.
• Designing and running red-team and purple-team exercises that simulate sophisticated adversaries and help validate the strength of existing defenses.
• Working with defensive security teams to convert offensive discoveries into concrete improvements, including stronger detections and better incident response readiness.
• Carrying out detailed penetration testing across the company’s product set, including AI-driven workflows, custom infrastructure, and user-facing interfaces.
• Developing and automating offensive tooling so testing can be scaled and adapted to emerging threats.
• Explaining findings clearly and persuasively to both technical specialists and non-technical partners so that risks are understood and mitigation work is prioritized.
• Helping influence security strategy by bringing attacker-informed insight into threat modeling, risk evaluation, and architectural decisions.

What the company is looking for

• Substantial offensive security experience, typically 7 or more years, or an equivalent track record that shows exceptional depth and achievement in the field.
• Strong familiarity with AI and machine learning security concerns, including prompt injection, data exposure, model manipulation, and abuse of dynamic interface components.
• Hands-on experience assessing modern technical environments such as custom Kubernetes and containerized deployments, cloud-native systems on AWS, GCP, or Azure, CI/CD pipelines, GitHub security practices, macOS and Linux internals, Python- and React-based applications, and data science or AI/ML tooling.
• A practical, builder-style mindset that includes writing robust tools, automating offensive workflows, and contributing to defensive solutions in complex codebases.
• Good instincts for identifying trust boundaries and evaluating risk in environments where priorities and systems change quickly.
• Excellent communication ability, including the skill to translate technical nuance into clear narratives that drive action.
• A collaborative approach and enthusiasm for working with engineers, researchers, and product teams to embed security throughout development.

Nice-to-have background

• Experience in AI, data science, or a related discipline, especially where security intersects with model behavior and system design.
• Experience in high-growth startups or research-oriented organizations where speed and innovation matter.
• Additional expertise in related areas such as software engineering, detection engineering, site reliability engineering, or security architecture.

Location and working style

The position is based in one of the company’s European offices, either Paris or London. Candidates are expected to already live there or be willing to relocate. The company places a strong value on in-person collaboration and encourages office attendance as much as possible, with a minimum expectation of three days per week on site. At the same time, the remote policy is intended to give employees flexibility, support work-life balance, and improve productivity.

What is offered

• Competitive salary together with equity in the form of stock options.
• Private health insurance.
• Support for transportation costs.
• A sports allowance.
• Meal vouchers.
• A private pension plan.
• Generous parental leave.
• Visa sponsorship.

How to apply

Applicants can submit their application through the job posting’s application link. By applying, candidates agree to the company’s Applicant Privacy Policy.

About the company

Mistral AI describes itself as a company focused on using AI to simplify work, save time, and support learning and creativity. Its platform is built to integrate into everyday work and is designed for enterprise use both on-premises and in cloud environments. The company emphasizes high-performance, optimized, open-source, and cutting-edge models, products, and solutions, and highlights its AI assistant, le Chat. It also describes its team as collaborative, low-ego, creative, and distributed across France, the USA, the UK, Germany, and Singapore.

Degree Requirement: Not Specified

Visa Sponsorship Promising