Can You Trust Your TPM?
Attacks on AMD Trusted Platform Modules raise security questions.
An article review.
Last year it was revealed that AMD Epyc and Ryzen CPUs were vulnerable to a series of attacks targeting their Secure Coprocessor–an entirely independent processor that can not be monitored or modified by the host machine. The Secure Coprocessor also hosts the Trusted Platform Module (TPM), which is used to store cryptographic keys used to secure the machine.
While the attacks were relatively minor and quickly mitigated with firmware updates, they raise questions about the secret computers that operate alongside our own devices. By necessity they must be independent from the host, but should an attacker gain access to these areas, they can operate completely in secret.
Now, we don’t think that we’re going to be able to get the tech companies to stop putting backdoors in our machines in the form of these secure enclaves, and they do provide meaningful security–when they haven’t been compromised. Secure or not though, if you’re not allowed to see inside how can you be certain about what’s going on?
Original article by Lorenzo Franceschi-Bicchierai writing for Vice.