Bypassing malware analysis sandboxes is easy, let’s discuss how they are doing it and why it works presented at BSidesSanFrancisco 2017

by Michael Gough,

Summary : Have you ever received a piece of malware and wanted to know what it did? You may have used an automated cloud malware analysis sandbox like VxStream/Reverse.It, Malwr or built your own Cuckoo sandbox. There are also high-end commercial solutions available such as ReversingLabs, Lastline and those integrated into Email, Web proxies and Next Gen Firewalls.
What about malware that uses documents such as .DOC and .PDF files versus a regular .EXE binary? Some sandboxes do multiple file formats, some do not and some claim to mimic user behavior. In comparing commodity malware, what we get in email and drive-by surfing to the advanced custom malware in targeted attacks I found some interesting things while doing manual and sandbox analysis. This talk will look at free, commercial and gateway (Email/Web) solutions and what can be learned from comparing the results to your own manual malware analysis.
Are these sandboxes worth it? Should you use them? What are the gaps? How much should you rely on the output? Do these solutions provide us what we need for Incident Response? Or enough data to improve the defense of our networks? Do they give us enough artifacts to remediate the infection?