
SAN FRANCISCO, Oct. 14, 2026 | OpenAI said Tuesday that it used its own AI models to identify an employee accused of hiding a listening device in one of its San Francisco offices, and that the FBI has arrested the person, according to CBC News. The charges are allegations from a criminal complaint and have not been proven in court.
What OpenAI and Prosecutors Say Happened
OpenAI found what it called a “sophisticated recording device that was hidden inside one of our offices” and reported it to the FBI, ABC7 San Francisco reported. The device was found on Sept. 30, hidden in a conference room inside toilet paper and plugged into a USB port on the back of a television. Federal prosecutors in San Francisco charged Abdulrahman Alfozan, who ABC7 describes as an OpenAI engineer, with using or attempting to use a device to intercept oral communications. He made his first court appearance on Tuesday, and it was not immediately clear whether he had a lawyer.
According to the complaint as reported by ABC7, surveillance footage showed him near the room about six times the previous month although he had no authority to enter it. After the company added a camera, he allegedly entered the room and checked behind the television. The complaint also alleges he used a productivity tool to make inquiries about building addresses, meeting rooms and calendars. Officials said the device was meant to capture sensitive conversations about the company’s technology, but the complaint does not say what he intended to do with any recordings. Alfozan should be presumed innocent unless and until he is found guilty, and his plea has not been reported.
How OpenAI Says It Used AI
In a post on X, OpenAI said it used its internal AI models to analyze the device and trace its origin, which pointed to a specific employee, Reuters reported. ABC7 reports that the FBI also helped identify the suspect through the added camera footage. OpenAI’s account of its AI use has not been independently verified, and the reports do not describe how the models were used.
OpenAI said it has “no indications” that the employee accessed or stole model weights, source code or customer data, and that he was terminated. The company said the investigation is ongoing. The FBI did not immediately respond to Reuters’ request for comment.
Frequently Asked Questions
Was any OpenAI data stolen?
OpenAI says it has no indications that model weights, source code or customer data were accessed or stolen. The investigation is ongoing.
Has the accused been convicted?
No. He has been charged in a federal complaint and made a first court appearance on Tuesday. His plea has not been reported, and he is presumed innocent.
