Back to the library

Key question New

AI-processed · Author-approved

Why do public employees resist mandated artificial intelligence tools when handling citizen data?

When Civil Servants Resist AI for Citizens: Privacy Concerns and Prosocial Resistance
Fynn E. Wojke · International Conference on Information Systems (ICIS) · 2026
Illustration for: Why do public employees resist mandated artificial intelligence tools when handling citizen data?

Loads Spotify's player; Spotify may set cookies. This device remembers your choice.

Key findings

  • Conceptualizes Custodial Boundary Turbulence to explain conflicts arising when administrative AI mandates breach a bureaucrat's protective obligations to citizens.
  • Posits that privacy concerns about AI misuse, specifically algorithmic bias and reduced human judgment, increase civil servants' resistance intentions.
  • Proposes that prosocial motivation amplifies AI privacy concerns and directly strengthens resistance intentions as an act of client advocacy.
  • Suggests that perceived workplace benefits and trust in organizational protective structures can attenuate resistance intentions and AI privacy concerns.

What it means for you

CIO / IT Executive

Establish and announce a formal 'Trust and Safeguards' framework on Monday morning that details how citizen data is protected within the new AI tool, addressing algorithmic bias and human oversight to build organizational trust.

IT Manager

Hold a Monday morning sync with frontline staff to explicitly map out where human-in-the-loop judgment points remain within the AI workflow, ensuring employees know they retain discretion over final decisions affecting vulnerable citizens.

Business Strategist

Audit the current change management strategy on Monday morning to replace top-down mandates with messaging that highlights how the AI system directly benefits citizens, aligning the rollout with civil servants' prosocial motivations.

Researcher

Draft a survey instrument on Monday morning operationalizing 'Custodial Boundary Turbulence' to empirically test how civil servants' prosocial motivation interacts with privacy concerns and resistance intentions among the 781 target employees.

Policymaker

Draft an addendum to the AI implementation mandate on Monday morning requiring explicit citizen consent protocols and transparent data-sharing boundaries to alleviate frontline workers' protective resistance.

Summary

This study examines why civil servants resist mandated artificial intelligence systems out of concern for citizen privacy. Drawing on Communication Privacy Management theory, the author introduces the concept of Custodial Boundary Turbulence and develops a research model to be tested with a cross-sectional survey of 781 non-managerial public administration employees.

Problem

Public sector mandates to implement automated AI tools can force civil servants to disclose sensitive citizen data without citizen consent or employee discretion. Existing research primarily views technology resistance through self-centered motives like status quo bias, overlooking protective, prosocial resistance enacted on behalf of vulnerable clients.

Transcript
Host: Welcome to A.I.S. Insights — powered by Living Knowledge. I'm Anna Ivy Summers.

Expert: And I'm Alex Ian Sutherland. Today, we are looking at a study titled "When Civil Servants Resist AI for Citizens: Privacy Concerns and Prosocial Resistance," authored by Fynn E. Wojke and published in the proceedings of the International Conference on Information Systems, or ICIS, 2026.

Host: In public administrations, leadership is actively introducing artificial intelligence to streamline processes. But this study examines an unexpected reaction: civil servants resisting mandated AI systems specifically to protect citizen privacy. Alex, what is the core problem here?

Expert: The central tension comes from how personal data is handled in the public sector. Unlike commercial platforms where users voluntarily hand over information, citizens typically have no choice. If you register a change of address or apply for social assistance, disclosure is legally required. The civil servant acts as an entrusted custodian of that data.

Host: And when leadership mandates an automated AI tool to process that data, that protective relationship is disrupted.

Expert: Exactly. Historically, information systems research explained workplace resistance through self-centered factors, such as status quo bias, where employees merely resist changes to their routine. This study points out that resistance can be prosocial and other-directed. Public servants often view themselves as client advocates, so when an AI tool risks compromising citizen privacy or fair treatment, they push back on behalf of the public.

Host: How does the author conceptualize this dynamic theoretically?

Expert: The author draws on Communication Privacy Management theory and introduces a concept termed Custodial Boundary Turbulence. Normally, privacy boundary coordination happens between the person who shares data and the recipient. But under an administrative mandate, a third party—the AI system—is introduced by institutional directive. When civil servants feel an ethical obligation to safeguard citizen data but lose the authority to control data flows, turbulence occurs. Because avoiding the issue is impossible under a mandate, that tension directly translates into resistance intentions.

Host: To investigate this, what empirical approach did the author design?

Expert: The study tests this model using a cross-sectional survey across five municipal administrations, serving populations ranging from roughly 13,000 to over 300,000 residents. The author retained a sample of 781 non-managerial employees who regularly handle sensitive citizen data across domains like social welfare, public health, and permitting. The analysis applies partial least squares structural equation modeling to evaluate the relationships.

Host: Let's discuss the key relationships in the model. What drives this resistance?

Expert: The framework examines privacy concerns about AI misuse, specifically focusing on algorithmic bias and reduced human judgment. When civil servants worry that automated tools will introduce biased decisions or remove critical human discretion, their intention to resist the mandate increases. Importantly, prosocial motivation strengthens these privacy concerns. Because these workers care about client welfare, they vicariously experience potential harms to citizens and elevate their protective boundary rules.

Host: Does the study identify factors that can reduce this resistance?

Expert: Yes. Resistance intentions are lower when employees perceive clear operational benefits from the AI tools. Furthermore, trust in organizational protective structures—such as institutional safeguards and privacy assurances—is modeled to attenuate privacy concerns. When workers trust that their administration has robust protections in place, their apprehension about AI misuse declines.

Host: What are the main practical takeaways for organizational leaders rolling out AI?

Expert: The practical lesson is that resistance should not be reflexively dismissed as mere inertia or technophobia. When frontline staff resist, it may reflect genuine benevolent stewardship toward the people they serve. Pushing harder on top-down mandates or focusing solely on efficiency arguments will likely fail. Instead, organizations need transparent protective structures, clear data safeguards, and governance models that preserve necessary professional discretion.

Host: Addressing those ethical boundaries seems critical for effective public-sector technology adoption. Thank you for breaking down the study, Alex, and thank you to our listeners for tuning in to A.I.S. Insights — powered by Living Knowledge.
Artificial Intelligence & Automation Information Systems & Management Human-Computer Interaction Data Analytics & Big Data Digital Transformation & Strategy Cybersecurity & Privacy Cloud & Infrastructure E-Commerce & Digital Platforms Operations & Supply Chain Sustainability & Green IS

Artificial Intelligence, Public Administration, Privacy Concerns, Communication Privacy Management, Prosocial Resistance, Algorithmic Governance