Offensive AI Security Engineer - Red Team

Job Title:Offensive AI Security Engineer – Red Team

We are seeking an Offensive AI Security Engineer to join our AI RedTeam within the Security Engineering team. This role focuses on adversarial machine learning (ML), AI-driven offensive security, and redteaming AI systems to uncover vulnerabilities in AI-powered automotive security models and vehicle platforms.

As part of Lucid's Offensive AI Security team, you will attack, manipulate, and exploit AI/ML models to identify real-world threats and weaknesses in AI-driven security solutions. You will develop AI-enhanced security automation tools, perform LLM-based penetration testing, and integrate AI/ML attack techniques into offensive security operations.

Key Responsibilities:

AI RedTeaming & Adversarial Attack Development

• Design and execute adversarial attacks on AI-powered security systems.
• Conduct LLM-based penetration testing to uncover AI security flaws in vehicle cybersecurity applications.
• Identify attack surfaces in AI-driven perception systems (LiDAR, Radar, Cameras) and develop exploits against automotive AI models.

AI-Driven Offensive Security Automation

• Build tools and systems to analyze emerging AI threats in vehicle security and enterprise AI models.
• Develop AI-assisted security automation tools for:
• • Reconnaissance – Automate vulnerability discovery using LLMs and RAG-based intelligence gathering.
  • Exploitation – Use AI to generate attack payloads and automate offensive security operations.
  • Fuzzing – Enhance automated fuzz testing with AI-driven input mutation strategies.
  • Reverse Engineering – Apply LLM-assisted binary analysis for rapid security assessments.
• Build offensive and defensive AI security tools, integrating ML-driven automation into security assessments and exploit development.

ML-Driven Security Research & Exploitation

• Use ML to characterize a program, identifying security-critical functions and behavioral anomalies.
• Understand LLVM Intermediate Representation (LLVM IR) to analyze compiled AI/ML software for security weaknesses.
• Develop AI-driven techniques for:
• • Threat Detection – Use ML to automate malware detection and anomaly recognition.
  • Cryptographic Algorithm Classification – Identify cryptographic weaknesses in compiled binaries.
  • Function Recognition – Use AI models to automate binary function analysis and decompilation.
  • Vulnerability Discovery – Automate zero-day discovery using ML-based exploit prediction models.
• Evaluate ML security models for robustness, performance, and adversarial resilience.

Offensive AI Research & RedTeaming Strategy

• Research novel AI attack techniques and evaluate their impact on vehicle cybersecurity and enterprise AI security models.
• Collaborate with internal RedTeams, SOC analysts, and AI security researchers to refine AI-driven offensive security approaches.
• Stay ahead of emerging AI threats, tracking advancements in AI security, adversarial ML, and autonomous vehicle AI exploitation.

Required Skills & Qualifications:

AI RedTeaming & Offensive Security Expertise

✔ Hands-on experience with AI/ML exploitation, adversarial ML, and AI-driven pentesting.

✔ Strong background in attacking AI models, including LLMs, deep learning systems, and computer vision AI.

✔ Experience with LLM-based vulnerability analysis, prompt engineering attacks, and model evasion techniques.

✔ Proficiency in penetration testing, AI fuzzing, and redteaming AI-driven security applications.

Cybersecurity & AI Security Research

✔ Experience in exploiting AI-powered vehicle security mechanisms.

✔ Ability to analyze, test, and exploit AI models within embedded systems and cloud environments.

Programming & Security Tooling

✔ Strong Python, C, C++ skills for AI security research and offensive security automation.

✔ Proficiency with AI security frameworks (TensorFlow, PyTorch, Hugging Face, OpenAI APIs, LangChain).

✔ Hands-on experience with reverse engineering tools (Ghidra, IDA Pro, Binary Ninja, Qiling, Angr).

✔ Familiarity with AI fuzzing tools (AFL++, Honggfuzz, Syzkaller) and adversarial ML frameworks (CleverHans, Foolbox, ART).

Preferred Qualifications:

• Bachelor's degree in Cybersecurity, AI/ML, Computer Science, or related technical field(s) is required. Master's degree or higher education is preferred.
• Experience in offensive AI security, redteaming AI models, or AI-powered security automation.
• Prior work in LLM-based security analysis, AI-driven pentesting, and adversarial ML research.