TSG Lab is a research group at the University of Oxford working at the intersection of AI safety and AI governance. We study how advanced AI systems represent information, how their behaviour can be evaluated and controlled, and how their development and deployment affect institutions and society.
The Lab is based in the Department of Engineering Science and works with the Oxford Martin AI Governance Initiative.

Independent research on AI safety and governance depends on sustained support. Funding helps us support researchers and students, provide computational resources, and produce work that informs both technical practice and public policy.