Abstract
<title>Abstract</title> <p>Digital anti-corruption tools—spanning big data analytics, AI-driven monitoring, and integrated digital platforms—are rapidly proliferating across the globe. Yet their effectiveness varies dramatically: some systems detect fraud at scale, while others degenerate into Potemkin surveillance. Existing research either assumes technological determinism (better technology equals less corruption) or falls into context-specific description without explaining cross-national variation. This paper asks: Why do digital anti-corruption tools that share similar technical architectures produce divergent outcomes across different political-institutional settings? Drawing on technological institutionalism and principal-agent theory, we develop a 'Technology-Institution Alignment' framework identifying three institutional mediators of digital anti-corruption effectiveness: (1) data integration capacity across government silos, (2) the autonomy and investigative authority of supervisory agencies, and (3) the legal-procedural legitimacy of algorithmic decision-making. Using a structured comparative case design, we analyze China's digital discipline inspection system alongside cases from Brazil, South Korea, and Estonia. Our findings suggest that digital tools amplify existing institutional capacities rather than substituting for them. The paper contributes to comparative public administration and digital governance scholarship by moving beyond descriptive accounts toward an explanatory framework, and offers policy guidance on the institutional preconditions for effective digital anti-corruption.</p>