Insiders and Immune Systems: When Rules and Signatures Don’t Work presented at CybersecurityCrucialCollaborations 2018

by Brad Hong, Dan Fein,

Summary : Rules and signatures struggle to keep pace with rapidly evolving cyber attacks. From insiders to sophisticated external attackers, the reality of cyber security often is that the threat is already inside. Specialists from the University of Cambridge developed a new approach to such threats, using unsupervised machine learning and probabilistic mathematics to establish an accurate understanding of normal behavior in an organization, across every network, device, and user; it spots abnormal activity as it emerges, and takes action to curb the threat.In this session, see real-world examples of unknown threats detected and contained through this technology, and explore its relevance to resource allocation and mitigation of risk.