AI Red Teamer/Penetration Tester - Remote

Overview

Mindgard is a London-based startup specializing in AI security. Our mission is to secure the future of AI Applications against cyber attacks. We’ve spun-out from a leading UK university after a decade of R&D, and are among the first few companies globally to offer solutions to this rapidly growing problem.

Mindgard’s automated AI red teaming product helps security testers test for, identify and remediate vulnerabilities in AI applications.

Mindgard probes AI applications and uses a set of techniques, including Predictive AI, LLMs, and other algorithms to identify security-relevant attributes within application responses.

We’re seeking an experienced AI Red Teamer who will be responsible for uncovering security vulnerabilities within AI. You will identify vulnerabilities that our customers care about, demonstrate successful cyber attacks, and report your findings to internal and external stakeholders.

In Short

• Red team and penetration test AI models, systems, and applications.
• Work with customers, AI providers, and the open-source community to disclose new security issues and help secure their AI solutions.
• Collaborate with engineering and research to design, implement, and report attacks against AI models and applications.
• Generate company awareness via regular engagement with the security community through publishing articles, releasing publicly available tools, and presentations.
• Stay up to date on the latest AI security research, trends, and attack techniques.
• Contribute to the development of internal tools, our AI red teaming product, and open-source projects.

Requirements

• Strong experience in red teaming and penetration testing.
• Hands-on experience and understanding of attack techniques specific to AI applications (prompt injection, jailbreaking, model extraction, etc.).
• Familiarity with AI red teaming tools and LLM vulnerability scanners (PyRit, Garak, etc.), as well as established pentesting tools (Burp Suite, Nessus, etc.).
• Able to communicate with technical and non-technical stakeholders.
• Excellent problem-solving skills and creativity to find and anticipate potential attack vectors against AI applications.

Benefits

• Inclusive work environment.
• Equal opportunities for everyone.