AI Red Teamer (Cybersecurity)

TL;DR

AI Red Teamer (Cybersecurity): Evaluating the security of predictive and generative AI models by simulating adversarial attacks and identifying vulnerabilities with an accent on adversarial machine learning and penetration testing. Focus on developing novel attack vectors like prompt injection and data poisoning to provide actionable security recommendations.

Location: Must be based in the United States (Fully Remote)

Company

Security solutions provider specializing in protecting Artificial Intelligence, agentic systems, and their underlying data.

What you will do

• Perform end-to-end penetration testing on both predictive and generative AI models.
• Execute adversarial attacks including evasion, poisoning, inference, jailbreaking, and prompt injection.
• Collaborate with data scientists and engineering teams to design and implement novel attack vectors.
• Produce detailed technical reports outlining vulnerabilities, risks, and mitigation strategies.
• Contribute to the development of internal tools and frameworks for AI red teaming operations.

Requirements

• 3+ years of experience in penetration testing, with at least 1 year focused on AI systems.
• Deep understanding of ML attack techniques such as prompt injection, jailbreaking, and model injection.
• Hands-on experience with AI security tools like Foolbox, CleverHans, ART, Purple Llama, or Garak.
• Proficiency in Python and experience with AI frameworks such as TensorFlow, PyTorch, or Keras.
• Strong ability to communicate complex security concepts to both technical and non-technical audiences.

Culture & Benefits

• Fully remote work environment with stipends for home office setup and monthly internet/phone expenses.
• Comprehensive health, vision, and dental insurance with 90% of premiums covered.
• 401k retirement offerings.
• Unlimited flexible time off in addition to 15 paid company holidays.
• Dedicated L&D fund for training, certifications, and industry conferences.