AI messaging scam costs Italy's top bank Intesa millions, sources say
Tap title to read story
Fraudsters employing advanced artificial intelligence voice-cloning technology successfully stole ninety-five million euros from Fideuram, the private banking subsidiary of Italy's premier lender Intesa Sanpaolo. The scheme began when executives received fraudulent communications purporting to be from top leadership requesting urgent overseas transactions. While international cooperation among authorities in Italy, China, and Portugal enabled the recovery of over half the stolen capital, approximately thirty-six million euros remains missing after conversion into cryptocurrencies. Fideuram detected the irregularities swiftly, triggering coordinated multi-jurisdictional investigations. Milan prosecutors have placed a foreign national under criminal investigation on suspicion of computer fraud, while former Fideuram Chairman Paolo Molesini resigned following the incident. This sophisticated cyberattack underscores escalating security vulnerabilities associated with generative artificial intelligence deepfake capabilities targeting global financial institutions, prompting heightened compliance and verification protocols across international banking networks to prevent similar high-value social engineering frauds.
More context
Timeline
- On February 1 2025 Fraudulent messaging initiated contact with Fideuram Chairman Paolo Molesini.
- On February 5 2025 AI voice cloning impersonated a prominent law firm partner.
- On February 10 2025 Unauthorized financial wire transfers were executed toward overseas accounts.
- On March 1 2025 Fideuram Chairman Paolo Molesini resigned citing personal reasons.
- On March 5 2025 Financial irregularities were detected prompting international bank alerts.
- On March 10 2025 Italian authorities launched official investigations into computer fraud.
- On April 1 2025 International cooperation recovered fifty-three million euros of funds.
- On May 1 2025 Investigators tracked remaining funds converted into untraceable cryptocurrencies.
- On June 1 2025 Prosecutors targeted a foreign national under criminal suspicion.
- On July 1 2026 Global financial institutions implemented enhanced AI verification protocols.
Why it matters
Immediate US impact: Financial institutions heighten internal verification protocols against AI scams. Possible long-term US impact: Accelerated adoption of cryptographic biometric authorization across commercial banking. Most affected groups: International banking executives, compliance officers, and cybersecurity professionals. Reader priority: Monitor enterprise authentication standards and emerging deepfake threat advisories.
Coverage notes
Left: Progressive outlets emphasize regulatory failures and corporate security vulnerability risks. Center: Neutral outlets report verified financial losses and official investigation updates. Right: Conservative outlets highlight international crime networks and lax cross-border controls.
Primary report
Reuters first reported the AI banking fraud involving Fideuram on March 21, 2025.
Sources
JQJO Editorial Desk · Human reviewed · AI assisted. Report a correction.
