Off The Record (unrecorded) "When ‘getting pwned’ doesn’t even fully describe what happened" When building your systems and infrastructure in the cloud, you should always consider the attack vectors that you open yourself up to and continually strive to proactively close them. It is common knowledge that when bringing up cloud computing resources you should implement controls such as preventing SSH logins as the root user, disabling password authentication for all users, and limiting which IP addresses can talk to the different services on your virtual machines, as well as requiring multifactor authentication for employees accessing cloud control panels. You can be fairly certain that an alarm would go off if an attacker was able to gain access, and that their access would be limited. But what happens if an attacker takes a different path and your infrastructure provider is compromised instead? Are your systems protected from that vector, and will your heuristics catch it? In this talk, we will tell a story, not from Netflix, but from the not too distant past around a successful targeted attack against a company using infrastructure providers as the vector. Details surrounding the methods used by the attacker will be shared, including the steps they took to attempt to cover their tracks. We will also look at how the attackers attempted to regain access after the initial vector was closed. Finally, we will look at what steps you can take to help mitigate the risks you incur if your infrastructure provider is compromised.