Postdoctoral Associate – AI Security
Core
Conducting original research at the intersection of machine learning, cybersecurity, and national security to secure advanced AI systems against sophisticated adversaries.
Role type
Postdoctoral Associate (Research Scientist)
Builds
Research findings, benchmarks, evaluation methodologies, and technical reports for government sponsors.
Domain
National Security / AI Security / Machine Learning
Deliverable
production ML models | research
Required skills
Adversarial machine learning, model robustness, secure AI system design, mechanistic interpretability, Python programming, ML frameworks (PyTorch, TensorFlow), empirical research design, interdisciplinary collaboration
Preferred skills
White-box threat models, MLOps, large-scale training infrastructure, AI system deployment architectures (RAG, multi-agent), adversarial evaluation frameworks, national security applications experience
Technologies
PyTorch, TensorFlow, Python
Responsibilities
Develop and evaluate novel attack and defense techniques for modern AI systems; Design and implement experimental frameworks for evaluating AI system vulnerabilities; Apply interpretability techniques to understand internal model behavior; Contribute to the development of benchmarks and evaluation methodologies; Translate research findings into actionable insights for government sponsors; Publish research in leading conferences and journals
Seniority
Postdoctoral (Early-career researcher)