This manuscript applies the Judgment-Centered AI Governance (JCAG) framework (Ayalew, 2026) to analyze Australia's Robodebt Scheme (2016–2020) as a paradigmatic case of algorithmic governance failure in public administration. The paper diagnoses three interrelated institutional failures that enabled an unlawful and harmful automated debt system to persist for over three years despite mounting evidence of its flaws, and argues that governance structures grounded in the JCAG framework's three pillars could have prevented the scheme's most damaging consequences. The analysis draws practical implications for AI governance in public sector contexts. The full manuscript is currently under peer review and access to the file is restricted.