Corporate employees and chief executives AI media incident
Federal prosecutors charged three individuals with a $129 million AI-driven fraud scheme that impersonated executives and lawyers to deceive employees into.
- Reported date
- Nov 7, 2026
- Target
- Corporate employees and chief executives
The exact incident date was not established. This entry is dated by its source report.
What happened
Federal prosecutors in New York unsealed an indictment charging three men with orchestrating a sophisticated, AI-enabled fraud scheme that targeted thousands of companies globally. Operating from at least July 2024 through February 2026, the group allegedly impersonated chief executives and lawyers from major firms to manipulate employees into authorizing unauthorized wire transfers. The scheme resulted in at least $129 million in documented losses, with victims including hospital systems, a British fintech company, and a mutual-fund firm.
The attackers utilized a multi-stage social engineering process rather than relying solely on technology. The process began with a legitimate-looking electronic-signature request, such as a DocuSign envelope, sent in the name of a company's chief executive. Once the employee signed a fake nondisclosure agreement, they were directed to a "compliance call" with a purported outside counsel. The attackers used AI to clone the voices and, in some instances, the faces of executives to add authority to their requests. The script emphasized extreme confidentiality and regulatory urgency, effectively isolating the employee from their colleagues and discouraging standard verification procedures.
In one instance involving a large New York investment bank, an employee was deceived into authorizing a $2.85 million transfer. However, the bank's back-office controls successfully blocked the wire, demonstrating that independent verification procedures—those not involved in the initial conversation—are the most effective defense against such attacks. The defendants, Jacob Soall, Sharon Hazan, and Guy Sheetrit, face seven counts, including wire fraud and aggravated identity theft, with potential statutory maximums totaling 104 years. The funds were laundered through nine jurisdictions and converted into cryptocurrency, complicating recovery efforts. This case highlights that while AI lowers the cost of impersonation, the success of these frauds relies on exploiting human trust and bypassing internal corporate verification protocols.