PhD Candidate in AI and Hardware Security

The Parallel Computing Systems group at the University of Amsterdam is seeking a motivated PhD candidate to research security vulnerabilities in AI systems. The project focuses on identifying and mitigating threats ranging from adversarial attacks and model extraction to physical side-channel and fault injection attacks on AI hardware. You will work across the algorithm, software, and hardware layers to develop robust, efficient, and practical security countermeasures. The role involves hands-on experimentation in a hardware lab, close collaboration with team members, and a commitment to academic excellence. You will also contribute to teaching and supervision duties within the Informatics Institute.

• Design and execute security evaluations on AI algorithms and hardware, including model extraction and side-channel analysis.
• Investigate the interaction of vulnerabilities across algorithm, software, and hardware layers.
• Develop and validate hardware-aware, robust, and low-overhead security countermeasures.
• Conduct experimental validation of methods on real hardware platforms such as NPUs and GPUs.
• Publish research findings in leading international conferences and journals.
• Complete a PhD thesis within the four-year contract period.
• Present research at international venues and foster academic and industrial collaborations.
• Contribute to teaching activities, including serving as a teaching assistant and supervising students.