Privacy of data processing is ensured using Fully Homomorphic Encryption (FHE), where an untrusted device computes on encrypted data. Integrity is assured through the use of Verifiable Computation (VC), where the untrusted device produces a proof of correct computation that can be verified by the user. Unfortunately, state-of-the-art approaches to FHE and VC incur a slowdown that ranges from four to six orders of magnitude over native computation, meaning that this approach is only viable for small-scale programs today. Srini Devadas describes ongoing research on hardware acceleration infrastructure to enable the deployment of FHE and VC for important classes of applications.