Contributions to research
Adversarial attack on ML systems is an emerging and relatively uncharted research territory for IS security and privacy research. Although IS scholars are using ML as a research method and studying ML systems as a source of superior profitability for firms, they have yet to theorize the emerging cybersecurity and privacy risks of ML systems. This study began to address this gap. The six constructs that serve as the conceptual building blocks of the research model in Figure 4 are likely to motivate IS scholars to use those constructs in future research. In addition, the nomological relationships among the six constructs, which are framed as testable propositions, are likely to foster empirical research on the harms caused to organizations by adversarial attacks and how to mitigate them.