Trusted execution environments for central banks

Trusted execution environments (TEEs) offer a hardware-anchored way to protect “data in use” by enforcing code authenticity, runtime integrity and confidentiality within a well-bounded trusted computing base. This paper considers what TEEs can responsibly deliver for central banks, the conditions under which they are most effective and how they can be deployed without widening who must be trusted. It takes a balanced view, outlining core assurances and boundaries, highlighting implementation and governance dependencies and situating TEEs alongside alternative approaches to confidential computing. The discussion links these foundations to central bank needs – such as cross-border data collaboration, supervisory analytics and operational resilience – and summarises practical considerations in performance, procurement, portability and auditability. While TEEs can enable deeper analysis and stronger resilience, their effectiveness is conditional and they should be embedded in disciplined engineering and evidence-based governance. The paper closes with a forward look at emerging directions and their implications for medium-term planning.