Italy · Rome · See the Italy profile · See the Rome profile
Top 62% 53/100 · Ask Evidence Copilot about this practice
Poste Italiane's 24/7 Fraud Prevention Center blocked about €25 million in fraud across 2 billion transactions in 2024 via ML risk scoring — but Italy's privacy regulator fined the group €12.5 million for requiring invasive device scans without a valid legal basis.
Poste Italiane, Italy's majority state-controlled postal and financial-services group, operates a 24/7 Fraud Prevention Center in Rome and Turin, staffed by more than 100 analysts, that shifted its digital-fraud detection from purely rules-based checks to machine-learning risk scoring.
In 2024 the center managed over 1 million risk alerts and blocked approximately €25 million in attempted fraud, reporting a 0.0012% fraud rate across 2 billion web-access attempts analysed and roughly 50,000 compromised accounts detected and blocked. It also identified more than 6,000 phishing 'clone' websites, with an average takedown time of three hours. On the insurance side, the center detected 66 fraud attempts among more than 800,000 settlement requests worth about €17 billion, blocking roughly €8 million.
A case study by the analytics vendor SAS, which names Poste Italiane's fraud-detection platform, reports that switching from rules-based detection to machine learning reduced false positives by about 40% and improved anomaly-detection capacity by more than 20%.
The same anti-fraud system drew scrutiny from Italy's data-protection authority (Garante), which found — following complaints filed in April–May 2024 — that Poste Italiane's BancoPosta and Postepay apps required users to authorise a fraud-detection tool built on LexisNexis ThreatMetrix to scan their devices (checking for other installed or running apps) as a condition of continued account access, without an adequate data-protection impact assessment or valid legal basis. In decision no. 237/2026 (17 April 2026), the Garante fined Poste Italiane €6,624,000 and Postepay €5,877,000, a combined €12.5 million. Poste Italiane announced it would contest the decision before the ordinary court in Rome, pointing to a separate ruling by the Lazio regional administrative court (TAR) on 2 February 2026 that had already annulled an earlier enforcement action against the same anti-fraud tool. According to the Garante's own website, the text of decision no. 237/2026 was later withdrawn from public access following a court order dated 17 July 2026 suspending its effects pending appeal — underscoring that the legal dispute over this system remains unresolved.
Read the full analysis: https://www.cybersecitalia.it/poste-italiane-con-il-fraud-prevention-center-nel-2024-bloccate-truffe-per-25-milioni/43640/
Implementation detail (cost, timeline, staffing, conditions for success) is not yet available for this practice.
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.
United States of America
USPS says AI tools helped close 1,250 fraudulent accounts since January 2026, but its own Inspector General found counterfeit postage …
Canada
Canada's signals-intelligence agency runs Assemblyline, an open-sourced ML malware triage platform, plus a separate classifier flagging threats commodity antivirus misses …
United States of America
The US Treasury's Office of Payment Integrity uses ML to screen federal disbursements. In FY2024, ML-driven processes prevented or recovered …
United Kingdom
The UK's Pensions Regulator and Pension Scams Action Group built a machine-learning tool to flag fraudulent pension websites: 830 sites …
Open full copilot Grounded in cited practices — always check the sources.