A fundamental security problem when hosting applications on cloud platforms is the increased risk of sensitive data loss (e.g. due to negligent or malicious employees of the cloud provider). An exciting approach to mitigating such attacks are new trusted execution environments (e.g. Intel SGX), recently available on commodity CPUs. Intel SGX allows users to create secure enclaves on remote cloud computers, such that enclave code and data is protected from an underlying malicious operating system or hypervisor, and also from physical attacks. Demand for these powerful security guarantees is evident from the rapid emergence of commercial SGX offerings in major cloud providers (e.g. Microsoft Azure Confidential Computing).
In the S3 lab, we are investigating several aspects of modern trusted execution environments, with a particular focus on Intel SGX. One issue is that buggy code inside enclaves may still leak sensitive data. A challenge here is program hardening techniques cannot rely on kernel support (e.g. for process isolation), since the OS is untrusted, increasing the importance of compiler-based hardening techniques. Another exciting line of research relates to potential risks of SGX-like technology. In particular, SGX raises the possibility of a new class of powerful malware that execute within enclaves, invisible to existing signature-based anti-virus tools.