Senior AI Safety Red Teamer

NVIDIA NVIDIA · Semiconductors · Tel Aviv, Israel +3

NVIDIA is seeking a Senior AI Safety Red Teamer to improve the safety and security posture of their AI models, systems, and infrastructure. The role involves hands-on safety and security research, developing tools to expose weaknesses, defining safety standards, and partnering with cross-functional teams. Requires 5+ years of experience in AI safety/security and offensive cybersecurity, with knowledge of AI vulnerabilities, LLMs, MLLMs, Generative AI, Agents, and RAG workflows, and strong Python programming skills.

What you'd actually do

  1. Drive hands-on safety and security research on a range of AI and networking products.
  2. Develop tools and processes to expose novel weaknesses in AI models and systems to preempt threats.
  3. Participate in defining and ensuring AI development processes meet safety and security standards.
  4. Partner with cross-functional teams to understand needs and implement solutions.
  5. Be a technical focal point across multiple teams and provide hands-on AI safety, security and engineering expertise.

Skills

Required

  • 5+ years of experience in AI safety/security and offensive cybersecurity
  • Knowledge of AI vulnerabilities and mitigation strategies
  • In-depth understanding of LLM, MLLM, Generative AI, Agents and RAG workflows
  • Python programming expertise

Nice to have

  • Hands-on experience designing and building software products, including infrastructure and system design
  • Knowledge of MLOps technologies such as Docker and Kubernetes
  • Familiarity with ML libraries (PyTorch, TensorRT, or TensorRT-LLM)

What the JD emphasized

  • hands-on safety and security experience
  • AI safety/security
  • offensive cybersecurity
  • AI (both model and infrastructure) vulnerabilities
  • LLM, MLLM, Generative AI, Agents and RAG workflows

Other signals

  • AI safety research
  • security research
  • AI red teaming
  • vulnerability assessment
  • risk reduction