My research vision is to develop learning-based methods that are robust against adversaries. We can achieve this by understanding the interplay program analysis, representations, and machine learning models play towards realizing Trustworthy AI for Systems Security.
I aim to establish the science of governing representations for trusted code reasoning: hierarchical, uncertainty-governed, cost-aware, semantics-grounded AI representations that make code analyses reliable to software change, adversaries, and capable of falsifiable guarantees and actionable explanations.
I am a Full Professor of Computer Science at University College London.