The integration of Automated Decision-Making (ADM) in the public sector promises efficiency but introduces critical risks of "authoritarian innovation." While existing literature explores algorithmic bias and citizen perceptions, the impact of these technologies on the illiberal behavior of civil servants remains under-researched. This proposal aims to investigate whether the technical authority of algorithms facilitates the adoption of illiberal practices by providing a structural "moral buffer" for public officials.

Drawing on Blame Avoidance Theory and Moral Disengagement, we argue that "protocolization"—the replacement of professional discretion with rigid algorithmic formulae—triggers a displacement of responsibility. By reducing job autonomy, ADM systems may allow bureaucrats to implement discriminatory or anti-democratic mandates while neutralizing the self-censure typically required to resist them.

As the empirical phase is ongoing, this presentation details our experimental design. We propose a survey-based experiment targeting civil servants to analyze how varying degrees of algorithmic mediation influence the willingness to execute illiberal instructions. By isolating displacement of responsibility as a causal mechanism, this research contributes to understanding how the "digital cage" of algorithmic governance can erode the internal institutional checks essential to liberal democracy.