OpenAI said it used its artificial intelligence models to identify an employee suspected of planting a concealed recording device in one of its offices, after which the company referred the matter to the FBI. The employee was arrested on 9 October and charged with using an electronic device to intercept oral communications, according to the US Department of Justice.
The employee, identified by the department as 35-year-old Saudi national Abdulrahman Alfozan, made his first appearance in federal court on Tuesday. The investigation remains ongoing, the department said.
In a post on X on Tuesday, OpenAI said it had discovered a “sophisticated recording device” hidden inside its San Francisco office and used its models to investigate its origin and identify the person allegedly responsible.
“We used our models to understand the device and where it had come from, identifying that a specific employee had planted it a few days earlier,” the company said. “We referred the matter to the FBI, and we’re grateful to the agents and prosecutors who moved quickly to make an arrest.”
According to the Justice Department, a criminal complaint alleged that an OpenAI employee discovered a USB device concealed behind a television in an executive conference room. The device had been hidden in toilet paper, the department said.
The complaint also alleged that closed-circuit television footage showed Alfozan near the conference room on several dates before the device was discovered. The allegations have not been established as findings of guilt in court.
The conference room was used for meetings involving highly confidential company information, according to the Justice Department. These discussions frequently covered proprietary research, product plans, business-development opportunities and non-public information about OpenAI’s technical progress, strategic decisions, capabilities and priorities. They also involved government engagements and commercial relationships, the department said.
OpenAI said it had found no evidence that the employee improperly accessed or manipulated the company’s model weights, source code or customer data. The company confirmed that the employee had been fired.
The precise role played by OpenAI’s AI models in identifying the employee remains unclear. The company did not immediately provide further details when approached for comment. It described the case as an example of how AI could help address “sophisticated security threats”.
The incident comes amid separate internal disciplinary action at OpenAI involving three employees working on safety and alignment. The company previously dismissed Jasmine Wang, Mikita Balesni and Tomek Korbak over alleged violations of its policies on handling sensitive information. OpenAI has said those dismissals were unrelated to the recording-device case.
The three former employees published a letter disputing some claims circulating about their conduct. They expressed concern about the lack of clarity surrounding their dismissals and warned that the way the company handled the matter could discourage transparency among staff.
OpenAI defended the decisions in an internal note, saying its investigation had uncovered misconduct beyond what was described in the former employees’ letter.
“Our internal investigation uncovered a significant breach of trust beyond what’s outlined in the letter they published and we stand by the decision to not continue their employment,” the company said.
