Fraudsters Allegedly Use AI-Generated Voice of Italian Defense Minister Guido Crosetto to Scam Business Leaders

In February 2025, scammers allegedly used AI voice cloning to impersonate Italian Defense Minister Guido Crosetto, calling top business leaders and reportedly claiming to need funds to free kidnapped journalists. Massimo Moratti was allegedly the only confirmed victim to send nearly €1M, wiring it to accounts in the Netherlands and Hong Kong before realizing the fraud. The scam involved spoofed government numbers, deepfake audio, and international money laundering networks. Police reportedly froze the funds.

Risk classification

• Primary risk domain: 4 Malicious actors
• Primary risk subdomain: 4.3 Fraud, scams, and targeted manipulation

The fraudsters used AI voice cloning to impersonate a government official and target specific wealthy individuals for financial gain, successfully stealing nearly 1 million euros from Massimo Moratti.

Causal factors

• Entity: Human
• Intent: Intentional
• Timing: Post-deployment

The incident was caused by human fraudsters who intentionally used deployed AI voice cloning technology to execute a financial scam.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Risk Level 3: Limited Risk. The incident involves AI-generated content such as deepfakes and voice cloning, which are subject to transparency obligations under the EU AI Act.

AI system and alleged parties

• AI system: unspecified AI voice cloning tool
• AI purpose: Voice Generation; unknown
• Behaviour type: Tool
• Alleged developer: Unknown voice cloning technology developers, Unknown deepfake technology developers
• Alleged deployer: Unknown scammers posing as Guido Crosetto, Unknown scammers
• Alleged harmed parties: Patrizio Bertelli, Massimo Moratti, Marco Tronchetti Provera, Italian business leaders, Guido Crosetto, Giorgio Armani, EssilorLuxottica, Esselunga, Del Vecchio family, Caprotti family, Caltagirone family, Beretta family, Aleotti family

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Substantial. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Substantial, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Negligible, indirect Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Financial

Reported: The reports explicitly describe a total financial loss of approximately 1,000,000 euros for the single successful scam occurrence, representing an average loss of 1,000,000 euros per occurrence.

Directly caused: Massimo Moratti transferred approximately 1,000,000 euros in two payments to a foreign account, which was the average financial loss per occurrence.

People affected

• Occurrences reported: 1
• People reportedly harmed: 1
• People reportedly exposed: 10

Potential causes

Management

• Outdated Security Policies: Corporate security guidelines did not account for advanced AI-driven threats.
• Inadequate Staff Training: Employees were not trained to recognize voice cloning and vishing tactics.

Technology

• AI Voice Cloning Technology: AI tools enabled highly convincing simulation of the minister's voice.
• Caller ID Spoofing Tools: Software allowed scammers to mask calls as official government numbers.
• Voice Cloning as a Service: VaaS platforms like PlugValley lowered barriers to deploy lifelike voices.

Data Inputs

• Public Voice Data Availability: Scammers used public recordings of the minister to train the voice clone.
• Target Profile Information: Detailed profiles of wealthy targets allowed precise selection of victims.

Human Factors

• Authority Bias: Victims trusted the request due to the minister's high-ranking status.
• Patriotic Duty and Urgency: Urgent pleas to save kidnapped journalists exploited victims' patriotism.
• Cognitive Overload: High-pressure scenarios made it difficult for victims to think critically.

Process and Methods

• No Multi-Factor Verification: Lack of verification protocols for high-value financial transactions.
• Lack of AI Detection Tools: Absence of real-time tools to identify cloned voices on active calls.
• Weak Call Screening: Some executive assistants failed to filter out suspicious foreign calls.

Regulatory Environment

• Weak Telecom Regulations: Telecom operators were not mandated to block spoofed caller IDs.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide consistent, detailed accounts of the scam, the victim, the amount of money involved, and the recovery process across multiple reputable news outlets. The role of AI voice cloning is explicitly stated by the Defense Minister and covered widely, leaving little ambiguity about the core events.
• Ambiguities identified: Some investigators initially considered simpler voice disguise techniques alongside AI voice cloning, though the consensus and the Minister's statements point to AI.
• Alternative interpretations: The incident could be interpreted primarily as a traditional social engineering and telecom spoofing scam where AI was merely an optional tool rather than the primary cause of the vulnerability.

Fraudsters used AI voice cloning to impersonate Italian Defense Minister Guido Crosetto in a sophisticated social engineering scam, successfully defrauding a business leader of nearly 1 million euros. While the direct national security impact is minor and the funds were frozen, the incident highlights the evolving threat of high-level government impersonation using synthetic media.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: Italy
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. The immediate fraud has been resolved and funds frozen, representing an ongoing strategic concern regarding AI-enabled social engineering.
• Autonomy: Human-controlled. The AI voice cloning system was directly controlled and utilized by human actors as a tool to execute the scam.
• Novelty: Evolved capability. AI voice cloning scams are established, but impersonating a defense minister to fabricate national security scenarios represents an evolved and highly targeted threat capability.

Impact by dimension

• Physical security: Negligible. No physical systems, critical infrastructure, or human safety were impacted by this financial scam.
• Information security: Minor. The incident involved AI voice cloning to impersonate the Italian Defense Minister, representing a minor intelligence and information security concern regarding leadership impersonation.
• Sovereignty: Minor. Fraudsters impersonated the Defense Minister and fabricated government actions (hostage ransom), but core government decision-making and sovereignty functions were not compromised.
• Economic security: Minor. While a prominent business leader was defrauded of nearly 1 million euros, the funds were frozen and there was no systemic threat to financial institutions or critical economic sectors.
• Societal stability: Negligible. There was no large-scale threat to social cohesion, civil liberties, or population safety.