XAVIER Mertens discusses statistics related to malicious PE (Portable Executable) files, focusing on compiler usage and extraction techniques. A Python script was developed to analyze PE headers and gather statistics from a dataset of 1.3 TB of malware samples. Key findings include that out of 23,501,548 files scanned, 690,689 were valid PE files, with a predominance of 32-bit binaries.
Statistics reveal high usage of various Microsoft Visual C++ linker versions, while tools employed by attackers were largely unidentified. The data shows a minimal presence of newer programming languages like Go and Rust in the datasets.