Distinguishing a human or machine cyberattacker

Abstract

In the world of cybersecurity, attacks may come from many sources. It is a problem faced by all users, because cyberattacks have become ubiquitous. Although many types of attacks are relatively simple to defend given commercially available detection software, the attacks are becoming more sophisticated. In more recent times, it is possible that a cyber attack might be initiated by a human or by an automated software attack or bot. One line of defense that has not been sufficiently explored is the advantage that might come to a defender if he or she has the ability to detect whether or not the attacker is human or machine-based. In this paper, we attempt to develop methods to allow a defender to try to determine the type of human or bot attack.