AI Security Engineer (m/f/n)

You will become part of the newly established AI Threat Research and AI Conversation Security team within the Technology Division. As a founding member, you have a unique opportunity to shape the team's direction and methodology from the ground up. The team researches attacks against conversational AI systems — probing and breaking them to understand how adversaries exploit large language models. Beyond research, the team designs and implements detection mechanisms and translates these defenses into shipping products across endpoint, cloud, and standalone form factors.

As a Conversational AI Security Engineer, you bridge adversarial research and production engineering — building detection mechanisms to defend conversational AI systems and shipping those defenses as products. Your work spans the full lifecycle, from breaking models to running detection in production.

About the Role

• Research attacks against conversational AI systems — exploring, probing, and breaking them to discover vulnerabilities before adversaries do.
• Design and propose mitigations, protections, and detection strategies based on findings from adversarial research.
• Build a multilayered detection core, combining different approaches to identify malicious prompts and attempts to manipulate AI systems.
• Productize detection capabilities — packaging and deploying them across endpoint, cloud, and standalone environments.
• Collaborate with the broader threat detection department, product teams, and endpoint teams to integrate defenses into shipping security solutions.
• Evaluate and experiment with open-source and proprietary AI tooling to stay current with the rapidly evolving attack surface.
• Shape the team's technical direction and methodology as a founding member, with the opportunity to grow into a Team Lead role over time.

Essential Requirements

• Strong software engineering background — ability to build production systems, not just prototypes.
• Proficiency in Python for building detection and security tooling.
• Understanding of how large language models work and how they can be manipulated or attacked, including prompt injection, jailbreaking, and adversarial inputs.
• Ability to work independently in a newly forming team with minimal defined processes.

Desirable Requirements

• Experience with machine learning and AI, including fine-tuning and deploying language models.
• C++ for performance-critical detection components.
• Background in cybersecurity or threat research.