Deploying Artificial Intelligence to Curb Corruption in the European Union
Irakli Petriashvili
2/8/20261 min read
This study presents the systematic map of artificial intelligence tools used to curb corruption across the European Union and offers quantitative evidence that countries adopting AI technologies experience meaningful reductions in corruption offences. Drawing on empirical data from Eurostat, the European Public Prosecutor’s Office, and the EU Public Sector Tech Watch, the research identifies forty-seven AI anti-corruption systems deployed by seventeen member states between 2016 and 2024.
Across these countries, police-recorded corruption offences fell by 17,592 cases. The author’s calculations indicate that financial cost of a per corruption case in EU member states range from 1.8 to 9.8 million euros. These figures translate into an estimated 31.7 to 172.4 billion euros in protected public value after AI anti-corruption tools implementation. A closer examination of the three highest AI adopters from 2016-2023 - demonstrates the scale of improvement: Italy achieved an 86.2 percent decline in corruption offences, while the Netherlands and Germany recorded reductions of 21.4 percent and 18.3 percent, respectively.
By contrast, Member States that did not deploy any anti-corruption AI systems, including Bulgaria, Croatia, and Romania, same period experienced persistent increases in corruption, in some cases exceeding 150 percent. Taken together, the findings provide the strongest cross-national evidence to date that AI-based systems can strengthen states’ capacity to reduce corruption. By producing the EU’s first AI anti-corruption map and demonstrating the scale of public resources potentially safeguarded through technological adoption, this study establishes a new empirical foundation for rethinking anti-corruption policy, guiding future investment decisions, and advancing the broader research agenda on AI governance and public sector integrity.
