Influence of motivational factors on hackers’ and analysts’ decisions in dynamic security games

Abstract

Cyber-attacks may be studied as a non-cooperative game between hackers and analysts. However, current game-theoretic approaches have disregarded how motivational factors (cost and benefit of hacker’s and analyst’s actions) are likely to influence decision-making during cyber-attacks. In an experiment, motivations of humans acting as hackers and analysts were manipulated across three between-subjects conditions in a repeated game: Equal-Payoff (Control; N = 25 pairs), Rewarding-Hacker (for successful attacks; N = 25 pairs) and Rewarding-Analyst (for correctly detecting cyber-attacks; N = 25 pairs). Hackers and analysts simultaneously decided in order to maximize their payoffs. Results revealed that the proportion of attacks was higher for Rewarding-Hacker condition and lower for Rewarding-Analyst condition compared to the Equal-Payoff condition. The proportion of defend actions was higher in Rewarding-Hacker condition and same in Rewarding-Analyst condition compared to the Equal-Payoff condition. We highlight the relevance of our results to cyber-attacks in the real world.