Research Scientist, AI and Systems Security - PhD New College Grad 2026

We are looking for a Research Scientist to advance security, safety, and privacy across three complementary research areas: (1) security of AI models and systems; (2) AI-enabled design and verification of secure computer systems; and (3) AI for cybersecurity. The first area includes making AI models, agents, data pipelines, runtimes, and accelerated computing platforms secure, safe, private, and resilient to attack, misuse, data leakage, and other failures. AI-enabled design and verification includes using AI, together with systems and formal techniques, to design, analyze, implement, and verify secure hardware and software systems. AI for cybersecurity includes applying AI to vulnerability discovery, secure systems development, threat detection, incident analysis, and automated defense. Across these three areas, we aim to build secure, safe, and private systems from first principles by co-designing hardware, software, AI models, algorithms, and operational controls. Our goal is to provide strong, measurable, and where possible verifiable protection while preserving performance, energy efficiency, and usability. We welcome researchers with depth in one area and interest in contributing across the broader agenda. NVIDIA is a leader in accelerated computing and AI systems. This role offers the opportunity to shape the security and safety foundations of emerging AI platforms, develop new approaches for AI-enabled secure system design and verification, and advance AI-powered cybersecurity in collaboration with research, product, and engineering teams.   What you'll be doing: • Advance the security, safety, and privacy of AI models and systems across the full lifecycle, including data collection, training, model and agent behavior, inference, deployment, and platform operations. • Develop AI-enabled methods to design, analyze, implement, and verify computer systems, including hardware architectures, system software, runtimes, compilers, and applications; rigorously evaluate security, safety, privacy, performance, and efficiency tradeoffs. • Advance AI for cybersecurity by developing techniques for vulnerability discovery, secure code generation and review, threat detection, incident analysis, security validation, and automated defense. • Collaborate across hardware, software, research, and product teams; publish original research; and help translate research advances into practical systems and developer capabilities.   What we need to see: • A Ph.D. in Computer Science, Computer Engineering, Electrical Engineering, or a related field, with a strong research background in security, AI/ML, computer architecture, operating systems, formal methods, programming languages and compilers, software systems, or other related areas. • A record of research excellence demonstrated through publications, open-source systems, patents, technology transfer, or substantial research collaborations. • Deep expertise in at least one of the three research areas—security of AI models and systems, AI-enabled design and verification of secure computer systems, or AI for cybersecurity—with the ability to connect that depth to challenges across the hardware/software stack. • Strong experimental research and prototyping skills, including the ability to build systems, define threat models, design meaningful evaluations, and analyze security and performance tradeoffs. • Strong programming skills in C, C++, Python, and/or relevant systems and machine-learning frameworks, along with clear written and verbal communication skills.   We are looking for creative, collaborative, and self-directed researchers who want to secure AI models and systems, use AI to help design and verify secure computer systems, and expand what AI can do for cybersecurity. Candidate