n an attempt to combat the issue of insider threats, the Department of Defense has tapped PARC, a Xerox company focused on innovation and R&D, to spearhead a new effort called the Graph Learning for Anomaly Detection using Psychological Context (GLAD-PC). The goal is to create technology that can automatically identify the possibility of a security threat coming from inside the department's network, leveraging large-scale behavioral data sets as well as information from social networks, among other sources, to determine when someone on the inside could pose a security risk.
Predicting insider threats by behavioral data before an incident happens? Sounds very similar to the movie Minority Report where police have created a system which predicts crime before it happens in a nightmarish Orwellian scenario.
The events surrounding Wiki Leaks over the past year coupled with other high profile insider attacks have fundamentally changed the way we approach security. Billions of dollars have been spent over the last few decades on IT security in order to “keep the bad guys out,” but it turns out the bigger threat was and always has been, found within the network perimeter.
Should warning signs of a potential malicious insider be addressed before a malicious event has occurred to prevent harm to the organization and discourage the insider from violating the organization’s rules? Predictive approaches cannot be validated a priori; false accusations may harm the career of the accused; and collection/monitoring of certain types of data may adversely affect employee morale.
While the GLAD-PC initiative could help to mitigate insider threats, we still believe the best method in preventing insider attacks is to implement a privileged identity management solution to create boundaries that enable end users and applications to communicate freely within an IT environment without worry of intentional, accidental or indirect misuse of privilege.
