You think your company's wire transfer protocol is secure? Think again. A massive federal indictment recently exposed how a trio of fraudsters weaponized artificial intelligence to impersonate corporate executives and high-powered attorneys, looting roughly $129 million from organizations worldwide.
This wasn’t a sloppy phishing email with broken English. This was a sophisticated, high-tech con job that blended deepfake audio, convincing video calls, and meticulous social engineering. Prosecutors in New York unsealed charges against Jacob Soall, Sharon Hazan, and Guy Sheetrit, revealing a multi-year cyber heist that should terrify every chief financial officer on the planet.
If you run a business or manage corporate finances, you need to understand how these criminals operated, why traditional safeguards failed, and what you can do right now to stop your company from becoming the next headline.
The Anatomy of a Multi-Million Dollar Deepfake Heist
Over a span of more than a year and a half, the accused syndicate targeted thousands of corporate targets across the globe. Their strategy relied on a terrifyingly simple formula: authority bias combined with bleeding-edge voice and video cloning.
The playbook usually started with compromised corporate email accounts or thorough reconnaissance. Once the fraudsters mapped out a target company's hierarchy, they would initiate contact pretending to be a CEO or top executive. When employees grew suspicious or asked for verification, the crew escalated.
They coordinated phone and video calls where artificial intelligence generated uncanny vocal replicas and visual imagery of the actual CEOs. Victims heard the familiar cadence, tone, and accent of their company leader telling them to comply with an urgent financial transaction.
To lock in the deception, the fake executives would direct employees to coordinate with a supposed outside attorney representing a major law firm. In reality, members of the syndicate—including Soall, who allegedly specialized in impersonating UK-based lawyers—jump on calls armed with a handwritten script titled "Lawyer Call" to sound authoritative and legalistic.
The targeted employees, convinced they were executing a confidential, high-stakes corporate maneuver under direct orders from top brass, wired millions out the door.
The Scale of the Damage
The numbers laid out in the federal indictment are staggering. This wasn't a pie-in-the-sky estimate; it represents millions siphoned directly from institutional accounts across various sectors:
- A hospital system in Missouri lost over $4.8 million.
- A bank headquartered in Rhode Island dropped nearly $2.7 million.
- A California hospital system parted with close to $3 million.
- A UK-based fintech firm handed over more than $11.5 million.
- A Massachusetts mutual fund and a Florida pet food company also fell victim, bleeding millions in individual strikes.
Once the funds hit external bank accounts, the conspirators quickly laundered the money. They converted the fiat currency into cryptocurrency to obscure the trail before distributing the cash amongst themselves.
Federal authorities eventually caught up with them across international borders. Soall was arrested in Spain in June, Hazan—who possessed a prior federal conviction for bank fraud in Manhattan—was picked up in Spain in August, and Sheetrit was nabbed in the United Kingdom in September. They now face severe federal counts of wire fraud and aggravated identity theft, carrying potential maximum sentences of 20 years per primary charge.
Why Traditional Corporate Controls Keep Failing
Most companies rely on dual-authorization policies and standard email verification for large wire transfers. Those defenses are entirely obsolete against real-time generative AI.
When an employee believes they are looking at their CEO's face on a video call and hearing their exact voice direct an emergency M&A transaction or regulatory settlement, policy often goes out the window. Fear of insubordination, combined with fabricated urgency, short-circuits rational caution.
Furthermore, criminals no longer need hours of pristine studio audio to clone a voice. Scammers can scrape public earnings call recordings, podcast appearances, social media videos, and promotional clips to train convincing deepfake models in minutes.
How to Protect Your Organization Today
You cannot stop bad actors from using AI, but you can build organizational friction that renders deepfake scams useless.
First, establish out-of-band verification protocols for any financial movement or sensitive data request. If a CEO requests an emergency wire transfer over a video call, the employee must hang up and contact that executive through a completely separate, pre-established internal channel—such as an internal messaging app or a direct callback to a known phone number.
Second, train your finance and accounting teams to spot the psychological triggers of fraud. Scammers always rely on urgency, secrecy, and emotional pressure. If an authority figure tells you to bypass normal compliance checks because of a strict deadline or a legal gag order, treat it as an automatic red flag.
Finally, implement zero-trust workflows for high-value transactions. No single person, regardless of how convincing the video feed or voice recording appears, should have the unilateral power to push millions of dollars across borders without multi-party physical or cryptographic confirmation.
The $129 million deepfake scandal is a watershed moment for corporate security. The technology to fake human presence is cheap, accessible, and improving every single day. If your company hasn't updated its security culture to account for synthetic media, you are running on borrowed time.
