GenAI Security Researcher
Core
Conduct adversarial research on frontier GenAI models and agents to uncover jailbreaks, prompt injections, and security vulnerabilities, translating findings into attacks, datasets, and evaluations.
Role type
Senior IC GenAI security researcher (offensive)
Builds
Adversarial datasets, attack taxonomies, evaluation scenarios, and automated attack generation/testing tooling
Domain
AI Security / Generative AI / Offensive Cybersecurity
Deliverable
production ML models
Required skills
Offensive security research, GenAI adversarial testing, LLM architecture understanding, Python automation, Threat modeling, Vulnerability research, Exploit development, Technical writing
Preferred skills
Published AI security research, AI bug bounty experience, Threat intelligence background, OWASP Top 10 for LLMs familiarity, Adversarial benchmarking, ML fundamentals
Technologies
Python, LLM-assisted tooling, Agent frameworks, MCP, RAG
Responsibilities
Conduct hands-on adversarial research on frontier GenAI models to uncover novel jailbreaks and guardrail bypasses; Research agentic attack surfaces including tool use and multi-agent pipelines; Assess model capabilities in offensive cyber domains; Translate classic offensive security techniques into attack scenarios against LLM-based systems; Design and build adversarial datasets and evaluation scenarios; Automate attack generation and testing using Python and agent frameworks; Communicate findings through reports, write-ups, and responsible disclosure.
Seniority
Senior, hands-on IC
