Verihubs Logo
HomeBlogWhen a Trusted Voice Becomes a Threat: The Growing Risk of AI Impersonation Fraud
6 min read•Published on November 19, 2026

When a Trusted Voice Becomes a Threat: The Growing Risk of AI Impersonation Fraud

When a Trusted Voice Becomes a Threat: The Growing Risk of AI Impersonation Fraud

Imagine receiving an urgent WhatsApp message from your company’s CEO requesting assistance with a financial transaction. Shortly afterward, a trusted legal advisor calls to confirm the instructions. The voice sounds familiar, and the request appears legitimate.

Would you question it?

For financial institutions, this is becoming a serious security concern. With artificial intelligence (AI), fraudsters can imitate familiar voices and identities, making fraudulent requests increasingly difficult to recognize.

A reported €95 million fraud involving Italian private bank Fideuram demonstrates how AI-powered impersonation can exploit trust and internal financial procedures.

The case raises an important question: How can businesses protect themselves when a familiar voice is no longer reliable proof of someone’s identity?

How an AI Impersonation Scam Led to a €95 Million Loss

In September 2026, Reuters reported that fraudsters had stolen approximately €95 million ($108 million) from Fideuram, the private banking arm of Intesa Sanpaolo, one of Italy’s largest banking groups.

The incident reportedly began in February 2026, when Fideuram’s then-chairman, Paolo Molesini, received a WhatsApp message that appeared to come from Intesa Sanpaolo CEO Carlo Messina.

The message requested urgent assistance with an overseas financial transaction.

Shortly afterward, a phone call reinforced the request. The caller appeared to be a senior partner from a law firm associated with the bank, using an AI-generated voice to imitate the lawyer.

Believing the instructions were genuine, the chairman authorized a series of transfers to overseas accounts, primarily in China and Hong Kong.

According to reporting by Biometric Update in October 2026, approximately €53 million was subsequently recovered through international cooperation.

However, a substantial amount remained unaccounted for.

The incident illustrates how fraudsters can combine familiar communication channels, executive impersonation, and AI-generated voices to influence high-value financial decisions.

Why Are AI Impersonation Scams So Convincing?

Executive impersonation fraud is not entirely new. Criminals have long used fraudulent emails and messages to pose as business leaders and request unauthorized payments.

However, AI technology adds another layer of credibility.

First, fraudsters exploit trusted identities.

Messages that appear to come from executives, legal advisors, or other senior professionals can create a sense of authority. Employees may feel pressure to respond quickly, particularly when requests involve confidential or urgent matters.

Second, AI-generated voices make impersonation more believable.

Voice cloning technology can produce speech resembling a real individual. This makes it harder to rely on familiar voices as evidence that a caller is genuine.

Third, attackers combine multiple communication channels.

In the Fideuram case, the reported WhatsApp message was followed by a phone call that reinforced the original request.

When two separate communications appear to confirm the same instruction, the request can seem more credible.

This combination of authority, urgency, and realistic impersonation makes AI-assisted fraud especially challenging for organizations that depend on rapid financial decision-making.

What Does This Mean for Banks and Financial Institutions?

The Fideuram case highlights a risk that extends beyond individual employees or executive accounts.

Financial institutions regularly handle sensitive information, high-value transactions, and communications involving multiple departments and external partners.

These activities often require employees to trust instructions from authorized individuals.

However, AI impersonation can make it difficult to establish whether a request genuinely comes from the person it appears to represent.

This creates several potential consequences, including unauthorized transfers, financial losses, operational disruption, and reputational damage.

The threat is also relevant to financial institutions across Southeast Asia, where digital banking, remote communication, and online financial services play important roles in business operations.

While strong authentication remains essential, organizations should avoid treating familiar voices, recognizable identities, or trusted messaging platforms as sufficient authorization for financial transactions.

The key lesson is that verifying a person’s identity must go beyond recognizing how they look or sound.

How Can Financial Institutions Prevent AI Impersonation Fraud?

Preventing AI-powered impersonation requires multiple security measures that address both human decision-making and digital verification.

1. Verify Sensitive Requests Through Independent Channels

Financial instructions involving unusual requests or significant amounts should always undergo independent confirmation.

For example, employees can verify a payment request by contacting the executive through an established company number rather than replying directly to the incoming message.

This helps reduce the risk of following instructions from impersonated individuals.

2. Strengthen Transaction Approval Procedures

High-value transfers should not depend entirely on a single communication or individual authorization.

Financial institutions can introduce multi-person approvals, transaction limits, and additional verification for unusual payment requests.

These controls can provide opportunities to identify suspicious activity before funds are transferred.

3. Educate Employees About AI-Based Impersonation

Security awareness programs should address how AI can be used to manipulate voices, images, and videos.

Employees need to recognize that even a convincing call from a familiar executive may not be authentic.

Regular fraud simulations and clear escalation procedures can help teams respond confidently to suspicious requests.

4. Strengthen Digital Identity Verification

Beyond executive impersonation, AI manipulation can also create risks during digital onboarding and remote customer verification.

Fraudsters may attempt to submit AI-generated facial images or manipulated videos to misrepresent their identities.

Businesses can reduce these risks through layered verification that combines identity matching, liveness checks, and deepfake analysis.

Each measure addresses a different threat, providing stronger protection when used together with transaction monitoring and risk-based authentication.

Strengthening Identity Verification with Verihubs Deepfake Detection

The Fideuram case shows how AI impersonation is changing the way businesses need to establish digital trust.

While preventing voice-cloning scams requires strong communication and transaction controls, businesses should also strengthen protection against manipulated facial identities.

Verihubs Deepfake Detection helps financial institutions identify potential AI-based facial manipulation through capabilities such as:

  • Facial Deepfake Detection: Analyzes images and videos for signs of AI-generated facial manipulation.
  • Fast Verification Results: Supports identity verification through automated analysis.
  • API and SDK Integration: Enables businesses to incorporate deepfake analysis into existing web and mobile verification workflows.

These capabilities can support digital onboarding, account verification, and other processes that rely on facial identity checks.

Deepfake Detection does not directly detect cloned voices or prevent fraudulent payment instructions, but it provides an additional security layer against AI-manipulated facial submissions.

Building Stronger Financial Security in the Age of AI

The reported €95 million Fideuram fraud demonstrates how AI-generated impersonation can turn ordinary business communications into opportunities for financial crime.

For banks and fintech companies, the lesson is clear: familiar identities and trusted communication channels are no longer enough to establish authenticity.

By combining independent transaction approvals, employee awareness, and stronger digital identity verification, businesses can build more resilient defenses against AI-driven fraud.

Learn how Verihubs Deepfake Detection helps businesses identify manipulated facial content and strengthen digital identity verification.

facebooktwitterlinkedinwhatsapp
Lihat Blog