In one of the most significant crackdowns on AI-enabled financial crime to date, federal prosecutors in the United States have unsealed charges against three individuals accused of orchestrating a sprawling $129 million fraud scheme. The operation, which targeted a diverse array of international corporations and healthcare systems, relied on the cutting-edge use of "deepfake" technology to impersonate high-ranking corporate executives and legal counsel, effectively tricking employees into authorizing massive, unauthorized wire transfers.
The defendants—identified as Jacob Soall, 41; Sharon Hazan, 54; and Guy Sheetrit, 42—allegedly leveraged artificial intelligence to bypass traditional corporate security protocols. By synthesizing the voices and likenesses of trusted leaders, the trio managed to siphon tens of millions of dollars from unsuspecting organizations, ranging from a Missouri-based hospital system to a prominent UK-based financial technology firm. The case highlights a terrifying new frontier in cybercrime, where the authenticity of a superior’s voice or a lawyer’s face can no longer be taken for granted.
According to the U.S. Department of Justice (DoJ), the conspiracy was built on a foundation of sophisticated social engineering, reinforced by AI tools that made the deception nearly indistinguishable from reality. The aim was simple yet devastating: to deceive staff members into transferring capital to a global network of bank accounts controlled by the conspirators. The scale of the theft, totaling more than $129 million, underscores the vulnerability of even the most established institutions to the evolving capabilities of generative AI.
The Anatomy of a Deepfake Heist
The methodology employed by Soall, Hazan, and Sheetrit was as calculated as it was technologically advanced. The DoJ’s investigation revealed a consistent pattern in how the fraudsters approached their targets. The scheme typically began with the impersonation of a company’s Chief Executive Officer (CEO). Using AI-generated voices or images, the fraudsters would contact employees—specifically targeting those within accounts departments or financial roles—and claim they required urgent assistance with a highly sensitive and "confidential" financial transaction.
The choice of targets was deliberate. By focusing on staff members who handled the movement of money but were far enough down the corporate ladder to feel pressured by a direct request from the CEO, the trio maximized their chances of success. The "confidential" nature of the requests served a dual purpose: it created a sense of executive-level urgency and provided a pretext for why the employee should not discuss the matter with their immediate supervisors or colleagues.
To further cement the legitimacy of the fraudulent requests, the perpetrators introduced a second layer of deception. After the initial contact from the "CEO," the targeted employee would be told to coordinate the details of the transaction with an attorney at a major international law firm. The fraudsters would then impersonate these lawyers, providing the specific wiring instructions and legal "justification" for the transfers.
The role of the legal counsel was primarily handled by Jacob Soall, a UK citizen. During the investigation, authorities discovered that Soall maintained a handwritten script titled “Lawyer Call.” This document was meticulously designed to manipulate victims, utilizing legal jargon and authoritative language to ensure the employee followed through with the wire transfer without seeking external verification. The script explicitly coached Soall on how to convince victims to maintain absolute secrecy regarding the communications, effectively isolating the employee within their own company. Once the employee was sufficiently convinced of the transaction’s legitimacy, the "lawyer" would provide the coordinates for a bank account where the funds were to be sent.
