AI Security Researcher

We are looking for a skilled AI Security Researcher to join our research team and work on advanced cybersecurity projects focused on emerging technologies, with a strong emphasis on AI and LLM-based systems.

You’ll research AI-based complex systems, identify and analyze vulnerabilities and attack vectors, develop PoCs, and build tools and methodologies to support hands-on security research. Our projects include software and security products, cloud environments, and advanced AI/LLM systems, including the layers that surround them such as orchestration, tool use, context management, safety mechanisms, and execution environments.

You’ll work alongside experienced researchers, take ownership of complex research missions, and contribute to high-impact projects for customers in Israel and internationally.

Responsibilites:

• Security & Vulnerability Research – Conduct hands-on security research across complex products, applications, systems, and technologies, with a focus on AI and LLM-based systems.
• Vulnerability Discovery & Analysis – Identify vulnerabilities and attack vectors in AI/LLM systems and their surrounding layers, analyze their root cause and security impact, and develop practical PoCs.
• Adversarial Research – Perform adversarial testing of LLM-based and agentic systems, including prompt injection, jailbreaks, tool misuse, data exfiltration, and excessive agency scenarios.
• Research Methodology – Explore unfamiliar technologies, map attack surfaces, define research approaches, and lead investigations from initial exploration through validated results.
• Technical Analysis – Perform analysis of AI/LLM systems (including black-box and grey-box approaches), code review of relevant components, and examination of orchestration, tool-use, context management, and execution environments.
• Tooling & Automation – Develop PoCs, evaluation harnesses, and automation to support research and enable repeatable testing of findings.
• Documentation & Collaboration – Document research methodology and findings, collaborate with other researchers and engineering teams, and present technical results internally and to customers.
• Ownership – Take end-to-end ownership of complex research tasks while working as part of an experienced multidisciplinary research team.
  
  

• 5+ years of hands-on experience in security research or offensive security.
• Proven hands-on experience in vulnerability discovery, security analysis, and PoC development.
• Strong programming and scripting experience, with strong proficiency in Python for building tools, automation, and research harnesses (C/C++ is a plus).
• Practical understanding of LLM architectures, tool calling, context management, and agent frameworks.
• Strong ability to independently research complex and unfamiliar systems and technologies.
• Ability to work within a team while leading complex research missions end-to-end.
  
  

• Hands-on experience in AI/LLM security research, including prompt injection, jailbreaks, or evaluation of LLM-based or agentic systems.
• Experience with low-level vulnerability research, exploit development, or fuzzing.
• Knowledge of embedded systems, firmware, OS internals, or communication protocols.
• Experience analyzing or building agent harnesses, orchestration layers, or tool integrations.
• Familiarity with AI security frameworks such as OWASP Top 10 for LLM Applications, OWASP Top 10 for Agentic Applications, or MITRE ATLAS.
• Experience with secure code review, security architecture review, or threat modeling of AI systems.
• Prior contributions to security research publications, CVEs, or responsible vulnerability disclosures (AI security or application security preferred).