Purportedly AI-Generated Calls Impersonating Banks Reportedly Helped Steal €380,000 from Customers in the Netherlands

Scammers allegedly used AI-generated bank-warning recordings to draw people in the Netherlands into calls with impersonators posing as bank staff. Victims were reportedly persuaded to install an app that relayed their bank cards' contactless data to accomplices at cash machines or payment terminals. Police reportedly documented about 170 victims and €380,000 in losses during three weeks of server monitoring.

Risk classification

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

The perpetrators used AI-generated voice clips to impersonate bank officials and deceive specific victims for illegitimate financial benefit.

Causal factors

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

The incident was caused by human criminals who intentionally deployed AI voice generation technology to facilitate a financial scam.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The AI system was used to generate synthetic audio/deepfake voices to impersonate bank employees. Under the EU AI Act, AI systems that generate or manipulate image, audio or video content (deepfakes) are classified as limited risk and are subject to transparency obligations.

AI system and alleged parties

• AI system: None named
• AI purpose: Voice Generation
• Behaviour type: Tool
• Alleged developer: Synthetic media technology developers, Synthetic audio generation technology developers
• Alleged deployer: Voice phishers, Synthetic media creators, Scammers, Deepfake creators, Cybercriminals in the Netherlands, Cybercriminals, Bank helpdesk fraud ring in the Netherlands
• Alleged harmed parties: Privacy, General public of the Netherlands, General public, Epistemic integrity, Banks, Bank customers in the Netherlands, Bank customers

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Minor. 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 Minor, 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 Minor, indirect Negligible
• Psychological: direct Negligible, indirect Minor
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Financial

Reported: The report explicitly describes a total financial loss of 380,000 euros across approximately 170 victims.

Directly caused: The average financial loss per occurrence was approximately 2,235 euros, totaling 380,000 euros across all victims.

Privacy

Reported: The report explicitly describes the unauthorized copying and transmission of bank card NFC chip data.

Directly caused: The malicious app scanned and transmitted NFC chip data from the victims' bank cards to the perpetrators' devices.

Inferred additional harm: The theft of bank card data represents a severe compromise of financial privacy and personal data security for the 170 victims.

Psychological

Reported: The report describes victims being put under pressure and manipulated, but does not explicitly quantify psychological harm.

Indirectly caused: Victims were subjected to high-pressure social engineering and manipulation during the fraudulent calls.

Inferred additional harm: It is highly likely that many of the 170 victims experienced significant distress, anxiety, and a loss of trust after being defrauded of thousands of euros.

People affected

• Occurrences reported: 1
• People reportedly harmed: 170
• People reportedly exposed: 170

Potential causes

Management

• Inadequate Public Awareness: Public campaigns were insufficient to warn about AI voice scams.
• Slow Initial Fraud Detection: Banks took weeks to identify and report the new NFC relay method.
• Lack of Bank Coordination: Banks and police had delayed response until the ECTF meeting.

Technology

• AI Voice Generation Tools: Allowed scammers to generate highly realistic and convincing voice recordings.
• NFC Relay App: Allowed remote transmission of card data to accomplices at ATMs.
• Automated IVR Spoofing: Used automated prompts ('press 1') to route victims to scammers.

Data Inputs

• Stolen Customer Contact Info: Scammers used targeted phone numbers to contact potential victims.
• NFC Chip Data Transmission: Unencrypted transmission of NFC card data over the rogue app.
• Bank Branding Elements: Use of specific bank names and scenarios to train or script AI voices.

Human Factors

• Susceptibility to Urgency: Victims acted under high stress and pressure to block their cards.
• Overtrust in AI Voice Realism: Victims believed the highly realistic AI voices were actual bank employees.
• Lack of Technical Literacy: Victims scanned physical cards against phones without knowing the risks.

Process and Methods

• Insecure App Installation Process: Sideloading of unverified apps allowed malicious NFC relay tools.
• Lack of Immediate Call Verification: No easy way for customers to instantly verify the validity of bank calls.
• Scripted Social Engineering: Structured scripts maximized psychological pressure on victims.

Regulatory Environment

• Lack of AI Voice Regulations: No strict controls on the creation and deployment of voice cloning tools.
• Permissive Sideloading Policies: Mobile operating systems allow installation of malicious non-store apps.
• Weak App Store Controls: Lack of prevention of external malicious apps accessing NFC hardware.

Information quality

• Classification confidence: High
• Reason for confidence: The reports from the Dutch police and AVROTROS are highly consistent, providing specific details on the number of victims (170), total financial damage (380,000 euros), the technical mechanism of the fraud (AI voice clips + NFC scanning app), and the arrest of the suspects. There are no major contradictions.
• Ambiguities identified: The specific AI models or platforms used to generate the voices are not named.
• Alternative interpretations: None. The reports clearly describe a coordinated cyber-enabled financial fraud campaign.

Dutch criminals used AI-generated voice cloning to impersonate bank officials, defrauding 170 victims of 380,000 euros. While representing an evolution in cybercrime capabilities, the incident was localized, managed by standard law enforcement, and posed negligible threats to national security.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: Netherlands
• Alleged perpetrator: Two Dutch citizens

Threat characteristics

• Imminence: Long-term. The specific operation was halted and suspects arrested, representing a long-term strategic concern regarding AI-enabled cybercrime rather than an active crisis.
• Autonomy: Human-controlled. The AI system functioned as a simple tool to generate static audio clips, with humans controlling the execution of the fraud and deployment of the clips.
• Novelty: Evolved capability. Represents an evolved capability where existing voice cloning technology is integrated into a multi-step bank fraud scheme.

Impact by dimension

• Physical security: Negligible. No kinetic threats, physical damage, or critical infrastructure compromise were reported in connection with this incident.
• Information security: Negligible. The incident involved localized financial fraud rather than state-sponsored information warfare, intelligence compromise, or mass surveillance.
• Sovereignty: Negligible. No threats to state authority, government operations, elections, or constitutional processes were identified.
• Economic security: Minor. The fraud caused a total loss of 380,000 euros. While targeting bank customers, it represents a minor, localized criminal threat managed by standard law enforcement.
• Societal stability: Negligible. The scam targeted individual bank accounts and did not involve mass surveillance, systemic discrimination, or threats to societal stability.