Alice — Brazil's AI system for detecting procurement fraud
Brazil
Alice, deployed by Brazil's CGU since 2015, uses NLP and RPA to screen federal procurement notices daily, reducing audit turnaround …
Italy · Rome · See the Italy profile · See the Rome profile
Evidence: Observational / pre–post Top 15% 80/100 · Ask Evidence Copilot about this practice
Italy's ANAC runs a public red-flag portal using ~70 corruption-risk indicators across 60m+ procurement records since 2022, flagging roughly one case a week and an estimated €935m/year in savings; 2026 added AI monitoring of PNRR tenders for anti-competitive clauses.
ITI4Italy's National Anti-Corruption Authority (ANAC) launched a public 'red-flag' portal on 21 July 2022, applying around 70 quantifiable, scientifically-selected indicators covering territorial context, procurement procedures and municipal-level risk to Italy's National Public Contracting Database (BDNCP), which has aggregated more than 60 million procurement procedures from over 39,000 contracting authorities since 2007.
Measure and flag corruption risk in public procurement using open data and standardised risk indicators, initially financed through the EU's 2014-2020 'Governance and Institutional Capacity' operational programme and later folded into Italy's PNRR anti-corruption measures.
Between 2017 and the 2024 review, ANAC ran 240 collaborative supervision projects with contracting authorities and issued more than 10,000 pre-litigation opinions on tender disputes; its data is published under the Open Contracting Data Standard (OCDS), queryable by API or downloadable as JSON. In August 2026, ANAC extended the programme to also flag anti-competitive clauses and serious regulatory violations in tenders funded by Italy's PNRR recovery plan, with initial focus on stalled Southern Italian construction sites, environmental criteria in road works, and school-building contracts worth EUR 3.24 billion.
A 2024 review by the Open Contracting Partnership found that the system's automated indicators help flag roughly one corruption case per week, and estimated the platform contributes 10-20% annual procurement savings in the health sector alone, approximately EUR 935 million per year.
The savings and case-detection figures are third-party estimates rather than a randomised or independently audited impact evaluation, and ANAC has not published the exact weighting of its ~70 risk indicators, so outside researchers cannot fully reproduce a given risk score. The approach nonetheless represents one of the most sustained, nationally-scaled uses of algorithmic risk-scoring for public integrity in the EU.
Recovery and Resilience Facility (national plans) National / regional programmes
Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.
Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.
Where this practice's information was retrieved from, and when.
Brazil
Alice, deployed by Brazil's CGU since 2015, uses NLP and RPA to screen federal procurement notices daily, reducing audit turnaround …
Bulgaria
Bulgaria's Ministry of Innovation and Digital Transformation launched SIGMA, a free, AI-assisted platform giving citizens real-time search over 193,019 public …
Kazakhstan
Kazakhstani anti-corruption NGO Adildik Joly uses Datanomix's Red Flags Management AI platform to flag risky public contracts, recovering KZT 244m …
Lithuania
Lithuania's STT, with the OECD and Government Transparency Institute (EU-funded), built a Hellinger Distance Random Forest on 11 merged datasets. …
Open full copilot Grounded in cited practices — always check the sources.