Three Arrested in Alleged AI-Assisted Executive Impersonation Fraud
Three men have been arrested in connection with an alleged fraud scheme that used artificial intelligence to alter and imitate the voices and images of corporate executives and attorneys. According to a press release from the U.S. Attorney’s Office for the Southern District of New York, the impersonations were intended to persuade employees at various companies to send large payments.
The men named in the announcement are Jacob Soall, Sharon Hazan and Guy Sheetrit. Soall and Hazan were arrested in Spain during the summer of 2026, while Sheetrit was arrested in the United Kingdom in September. The alleged activity took place between 2024 and 2026.
The available account does not identify the companies targeted, specify how much money was sought or transferred, or describe the particular AI systems used. It also does not provide details about the alleged communications or the legal charges. Those points should not be inferred from the arrests alone: the allegations remain allegations, and an arrest is not a determination of guilt.
Why the reported method matters
The reported scheme combines familiar payment fraud with synthetic representations of trusted people. A request that appears to come from an executive or legal adviser may carry authority inside an organization, especially when it concerns a large or time-sensitive transfer. The alleged use of manipulated voices and images adds another layer to impersonation, but does not make those representations reliable proof of identity.
For enterprise teams, the practical lesson is to keep payment approval separate from the communication channel used to request it. A voice call, video interaction or image should not replace established authorization procedures. Employees should verify unusual or high-value payment instructions through a previously known contact method, rather than relying on contact details or links supplied with the request.
Controls for payment and identity verification
Organizations can review whether their payment workflows require independent approval, documented business justification and confirmation through a second channel. Those controls are useful against impersonation whether a fraudster uses conventional social engineering or AI-altered media. Finance, legal and executive-support teams should know how to pause a request and escalate it without being pressured to bypass normal procedures.
Training should also address the limits of audio and video as identity signals. Staff can be asked to treat unexpected requests for large payments as verification events, even when the apparent sender is familiar or the communication looks convincing. Internal guidance should make the approved verification route easy to find and should explain how to report suspected impersonation promptly.
The arrests span more than one country, but the announcement described here does not provide further details about international coordination or the investigation. It does establish the central alleged tactic: using AI-altered voices and images to pose as company executives and attorneys in pursuit of large payments. Enterprises can respond without assuming any particular tool or scenario by reinforcing payment controls, validating identity independently and ensuring employees are not left to make high-stakes decisions based on media alone.
