Living-off-the-land (LotL) is used to describe attacks where malicious actors leverage legitimate software in an effort to avoid being detected.

Adobe has released an open source tool, named LotL Classifier, that is designed to detect LotL attacks by leveraging a “feature extraction” component and a machine learning-based classifier algorithm.

The feature extraction component takes data from threat intelligence, malware analysis, real incidents and real data logs, and uses that data to generate a series of tags based on binaries, paths, keywords, networks, patterns, and similarity.

The tags are then fed to the classifier component, which decides if the analyzed data set is good or bad. This component also creates a set of tags that can be integrated with rule-based automation or anomaly detection tools, such as One Stop Anomaly Shop (OSAS), which Adobe recently released as open source