Federal prosecutors in Manhattan have charged three men with using artificial intelligence to impersonate company chief executives and lawyers, in a fraud scheme that prosecutors say tricked employees into wiring away more than $129 million. The indictment, unsealed on Thursday in the Southern District of New York, accuses the men of cloning the voices — and in some cases the faces — of senior figures at businesses and major law firms, then using those convincing imitations to order large payments to bank accounts they controlled.
The three defendants are Jacob Soall, 41, a Spanish national who lived in Marbella in Spain and in the United Kingdom; Sharon Hazan, 54, a British and Israeli citizen; and Guy Sheetrit, 42, an Israeli citizen who also holds Emirati nationality and lives in the United Arab Emirates. Each faces seven federal counts, including conspiracy to commit wire fraud, wire fraud, aggravated identity theft and money laundering. The fraud and money laundering counts each carry a maximum sentence of 20 years in prison, and prosecutors say the combined maximum across the charges is 104 years.
According to the indictment, the scheme ran from July 2024 to February 2026 and followed a script that investigators say was designed to exploit trust inside a company. The group would first identify an employee who was authorised to transfer large sums of company money. Then, posing as that employee's own chief executive, they would make contact — by phone or video call — and ask for help with what was presented as a confidential financial transaction. The voice on the other end of the line, and sometimes the face on the screen, was not the boss at all, but an AI-generated imitation.
Prosecutors say the deception worked precisely because the employees believed they were speaking to their own leadership. Once that trust was established, the supposed executive would direct the employee to wire a large sum to a particular bank account. Only later would the company discover that the account belonged to members of the conspiracy, and that the money — in some cases millions of dollars from a single transfer — was gone.
The indictment names a series of victims across several countries and industries. A financial technology company headquartered in the United Kingdom lost $11,578,474, the largest single loss listed. A pet food maker in Florida lost $5,242,747, a hospital system in Missouri lost $4,829,580, and a mutual fund company in Massachusetts lost $4,639,918. A hospital system in California lost $2,968,497 and a bank headquartered in Rhode Island lost $2,684,580. Prosecutors say there are further victims still being identified, and the FBI has urged any business that believes it was targeted to come forward through its online tips page.
All three men were arrested in Europe in recent months and are awaiting extradition to the United States. Spanish police detained Soall on 17 June and Hazan on 5 August, while Britain's National Crime Agency arrested Sheetrit on 7 September. Court records listing defence lawyers for the men were still sealed on Thursday evening, and the charges remain allegations: all three defendants are presumed innocent unless proven guilty in court.
One of the defendants, Hazan, has a history with the Manhattan federal court. Prosecutors note that he was previously convicted there of conspiracy to commit bank fraud in 2008. Reports of the case also say that one suspect boasted of having laundered $400 million, while another flaunted luxury cars and a computer said to be worth $200,000 — details that, if proven, would point to a group that treated the proceeds of the scheme as a lifestyle rather than a secret to be hidden.
The scale of the alleged fraud sets it apart from earlier deepfake scams. The $129 million haul is roughly five times the $25.6 million that British engineering firm Arup lost in 2024, when an employee in Hong Kong joined a video call with a deepfake of the company's finance chief. It also dwarfs the just over $30 million in AI-linked business email compromise losses that victims reported to the FBI for the whole of 2025 — a figure the bureau counts only when victims themselves recognise that AI was used, which suggests the true scale of the problem may be larger still.
Announcing the charges, US Attorney Jamie McDonald said the technology behind the scheme "offers tremendous potential. But the public should recognise that there are some who seek to use this emerging technology to pursue criminal goals." James C. Barnacle Jr., the assistant director in charge of the FBI's New York field office, said the case showed how criminals were weaponising emerging technologies, and that the bureau was adapting its own capabilities to identify and disrupt them.
The case is now expected to move through extradition proceedings in Spain and the United Kingdom before any trial in Manhattan, a process that can take months. In the meantime, investigators say they are still working to identify additional victims, and security experts say the episode is a warning to companies everywhere: payment requests, however senior the voice making them, should be verified through a second channel before any money moves. As AI imitation tools grow cheaper and more convincing, the line between a genuine instruction from a boss and a fabricated one is becoming harder to hear — and, prosecutors argue, this case shows what happens when that line is crossed.
