Scoring files with Endpoint Defense AI

The Aurora Engine uses Endpoint Defense AI to classify files as bad or good with a certain level of confidence. The classification is made based on machine learning models. The models are being constantly improved using machine learning to refine the accuracy of the results.

The models contained in this version of the Aurora Engine are listed in the table below. Each of these models requires a corresponding dynamic library (a shared object in Linux terms).

File type

Model name

 Extensions

Windows executables

Ensemble-20230818-S3V3-PE7E.cym

.acm, .ax, .cpl, .drv, .efi, .mui, .ocx, .src,

.sys, .tsp, .exe, .dll

macOS executables

Ensemble-20210721-S3V4-MO3.cym

Ensemble-20210409-S0V2-MOFAT.cym

(none), .o, .dylib, .bundle
Linux executables

Ensemble-20180730-S2V7-ELF2.cym

(none), .o, .ko, .mod, .so

OLE files

Ensemble-20180718-S3V3-OLE3.cym

.doc, .xls, .ppt

OOXML files

Ensemble-20180718-S3V3-OOXML3.cym

.docx, .xlsx, .pptx

PDF files

Ensemble-20230607-S3-PDF4.cym

.pdf

Archive files

Ensemble-20190319-S0V5-ARC.cym

.zip, .7z, .rar, .tar, .gz, .bz2, .xz

Note: Files with the .zip extension are supported whether they are password-protected or not. Files with the .rar extension are supported only if they are not password-protected.