Ransomware and the Robin Hood effect?: Experimental evidence on Americans’ willingness to support cyber-extortion

Abstract

Objectives Ransomware attacks have become a critical security threat worldwide. However, existing research on ransomware has largely ignored public opinion. This initial study identifies patterns in the American public's support for the use of ransomware, specifically when it is framed to provide benefits to others (i.e., in-group members). Drawing on the Robin Hood decision-making literature and Moral Foundations Theory, we offer theoretical predictions regarding ransomware support. Methods In a survey of 1013 Americans, we embedded a split-ballot experiment in which respondents were randomly assigned to indicate their level of support or opposition to one of two sets of six ransomware scenarios. We manipulated the nationality, authority level, and political affiliation of the actors. ResultsWe find that people are more supportive of ransomware use when the actors are from their own in-group, and the outcomes benefit their in-group members. Also, the more strongly participants endorsed the moral foundations of authority and harm/care, the more supportive they were of the use of ransomware that may benefit others from their in-group. Conclusions These findings suggest political actors may be able to generate public support for morally questionable actions by emphasizing in-group benefits and the Robin Hood nature of an attack (e.g., outcome-based morality).

Did George Floyd’s murder shape the public’s felt obligation to obey the police?

“…Participants were recruited through Prolific, an online survey platform that samples users on the basis of researcher criteria (Barrowcliffe et al, 2022; Turner et al, 2021). Platforms such as Prolific use opt-in processes to yield convenience samples that are typically more diverse than college student samples and less prone to social desirability biases (see Haner et al, 2022; Pickett et al, 2022; Pickett, Roche, & Pogarsky, 2018). Prolific also uses an intensive process to prevent botlike accounts from completing surveys to enhance the quality of the data (Bradley, 2018).…”

A Survey on Ransomware Threats: Contrasting Static and Dynamic Analysis Methods

“…The integration of these approaches, particularly when underpinned by machine learning models trained on a comprehensive spectrum of ransomware behaviors, holds promise [11,44,69]. Such models, informed by a broad dataset reflective of the diverse tactics employed by ransomware developers, could provide a more robust defense against the broad array of ransomware attacks witnessed today [51,70,71]. The development and training of these models require a meticulous approach, accounting for the ever-changing tactics that ransomware developers employ to avoid detection [37,41,72].…”