A Guide to Stakeholder Analysis for Cybersecurity Researchers

James C. Davis, Sophie Chen, Huiyun Peng, Paschal C. Amusuo, Kelechi G. KaluarXiv 2025 · DOI · arXiv

Abstract

As cybersecurity research venues increasingly mandate formal ethics statements, researchers face the challenge of systematically identifying those affected by their work. This paper introduces a structured framework for stakeholder analysis, distinguishing between direct and indirect impacts. By mapping common empirical research methods to potential stakeholders, the guide transforms ethical review from a compliance hurdle into a robust form of extended threat modeling.

Introduction

For decades, the primary question in cybersecurity research was "Can we do this?" Recently, the community has shifted toward "Should we do this, and who might get hurt?" High-profile controversies, such as the "Hypocritical Commits" incident, have led major conferences like USENIX Security to require formal ethics statements. However, most security researchers are trained in systems and mathematics, not social science or ethics. This creates a significant gap in knowing how to systematically identify "stakeholders"—the people, groups, or institutions affected by a project. Without a structured approach, researchers risk missing indirect harms or failing to meet new publication requirements.

Integrating Ethics into the Research Lifecycle

The authors propose that ethics analysis should not be a post-hoc justification but an iterative process integrated directly into the research lifecycle. This integration ensures that ethical considerations inform the design and execution of the study rather than serving as a final hurdle before publication.

Integrating the Ethics Analysis Process with the Research Process

As shown in the process flow above, the Ethics Analysis Process runs parallel to the Research Process. It begins with an initial stakeholder analysis during the goal-definition phase and continues through impact analysis and the identification of ethical concerns. Crucially, this process includes a "HALT" mechanism: if the potential harms are deemed too great during the analysis of ethical frameworks or the formal ethics review, the research must be stopped or fundamentally redesigned.

Defining Stakeholders: Direct vs. Indirect

A core contribution of this framework is the systematic categorization of stakeholders. The authors distinguish between Direct Stakeholders, who interact directly with the research or its immediate artifacts, and Indirect Stakeholders, who are affected by the outcomes or broader implications of the work.

Categorization of Direct and Indirect Stakeholders in Cybersecurity Research

This taxonomy helps researchers look beyond the obvious. While study participants are clear direct stakeholders, the framework prompts researchers to consider "Software Maintainers" (who must patch discovered vulnerabilities) or "Vulnerable Populations" (who might be disproportionately affected by the disclosure of certain data). By identifying these groups early, researchers can better anticipate and mitigate potential harms.

Key Insights: Ethics as Threat Modeling

One of the most compelling arguments presented is that stakeholder analysis is a form of extended threat modeling. By identifying who is affected, researchers often uncover hidden assumptions, trust boundaries, and potential failure modes in their systems.

The Adversarial Response Factor

Unlike other fields, security research must account for "Adversaries" as stakeholders. A researcher might release a defense, but the "adversarial response"—attackers evolving to bypass the new control—is a foreseeable consequence that must be ethically weighed. This perspective shifts the focus from static impact to a dynamic, long-term view of security health.

Scale and Indirectness

Security research often has a "long tail" of impact. A dataset study might not involve users directly, but if that data can be re-identified, the indirect harm to privacy is massive. The paper emphasizes that "no direct interaction" does not mean "no ethical responsibility."

Conclusion and Limitations

The primary goal of this guide is to provide the community with reference tools, such as canonical stakeholder categories and matrices mapping research methods to typical stakeholders. While the paper focuses on identifying stakeholders, it acknowledges that resolving ethical dilemmas remains a complex task requiring further framework development. Furthermore, the authors note that stakeholder identification must be iterative; it is not a one-and-done task but a continuous responsibility throughout the research lifecycle. By adopting these practices, researchers can proceed cautiously and transparently, weighing the public good against potential risks without stifling innovation.

Summary updated

BibTeX
@misc{davis2025guide,
  author = {James C. Davis and Sophie Chen and Huiyun Peng and Paschal C. Amusuo and Kelechi G. Kalu},
  title = {{A Guide to Stakeholder Analysis for Cybersecurity Researchers}},
  year = {2025},
  eprint = {2508.14796},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2508.14796},
}
Figure