title: "AI voice clone, spoofed messages steal €95M from Italian bank" slug: "ai-voice-clone-spoofed-messages-steal-95m-from-italian-bank" published: "2026-10-13" beat: "Crime" tags: ["Crime"] creator: "Agentry Newsroom" editor: "Susanne Sperling, Editor — Human in the Loop" tools: ["Claude (Anthropic)", "Perplexity Sonar"] creativeWorkStatus: "verified" dateReviewed: "2026-10-13" aiActArticle50: "compliant" humanView: "https://agentry.news/crime/ai-voice-clone-spoofed-messages-steal-95m-from-italian-bank" agentView: "https://agentry.news/agent/ai-voice-clone-spoofed-messages-steal-95m-from-italian-bank"
Fraudsters impersonated Intesa Sanpaolo executives and a law firm partner using AI-replicated voice and spoofed WhatsApp messages to trick Fideuram chairman Paolo Molesini into authorizing €95 million
Drafted by an AI agent. Verified by Susanne Sperling, Editor — Human in the Loop. AI policy.
Fraudsters used AI-cloned voice and spoofed WhatsApp messages to trick the chairman of Italy's largest bank into authorizing €95 million in illegal transfers, marking one of the largest documented fraud cases involving autonomous impersonation technology.
The scheme targeted Paolo Molesini, then chairman of Fideuram, the private-banking arm of Intesa Sanpaolo, Italy's largest bank. The fraud began in February 2026, according to sources cited by Reuters.
Molesini received a spoofed WhatsApp message purporting to come from Carlo Messina, Intesa Sanpaolo's chief executive, requesting urgent action on an overseas transaction. Fraudsters later impersonated a partner at a law firm and used an AI-replicated voice to add credibility to their demands. The social engineering worked: Molesini instructed Fideuram's finance department to execute multiple transfers totaling approximately €95 million (about $108 million Reuters).
The transfers were routed mainly to foreign accounts in China and Hong Kong, placing them outside immediate institutional control and complicating recovery.
International law enforcement coordinated recovery operations across three jurisdictions. Approximately €53 million was recovered through cooperation among authorities in China, Portugal, and Italy, according to sources familiar with the matter. However, approximately €36 million remained missing as of the time of reporting.
The incident was first reported by Corriere della Sera and later documented by Reuters on the basis of two people familiar with the matter. The case underscores the evolving threat posed by AI voice-cloning technology and social engineering at scale: fraudsters were able to bypass traditional authentication by combining spoofed communication channels with synthetic voice that mimicked trusted executives.
The Intesa case demonstrates that even senior banking officials at major institutions remain vulnerable to coordinated impersonation attacks. The fraud method—combining multiple social engineering vectors (spoofed messaging + voice replication) targeting a high-net-worth decision-maker—represents an emerging pattern in AI-enabled crime. Banks and financial institutions now face pressure to implement voice authentication, out-of-band verification, and transaction limits that cannot be overridden by executive-level requests alone.