Abdulrahman Alfozan, a former product engineer at OpenAI, has been arrested and charged with allegedly planting a hidden recording device inside the company’s San Francisco headquarters.
The device was reportedly tucked into a high-security executive conference room.
The FBI charged Alfozan on October 9, 2026, with allegedly using an electronic device to intercept oral communications. OpenAI terminated his employment the same day.
How the alleged scheme unraveled
OpenAI discovered a small audio recording device on September 30, 2026, during a routine check of the executive conference room. Investigators believe it had been placed there approximately two days earlier, around September 28.
On October 5, 2026, a second, similar device was found in Alfozan’s personal bag.
Security footage also reportedly captured Alfozan near the conference room on multiple occasions. His job did not grant him permission to enter that sensitive area.
Investigators determined the device was designed to capture confidential discussions at the company.
OpenAI used its own internal AI models to help trace who was responsible for the device.
A one-way ticket and a long résumé
Alfozan had booked a one-way flight to Saudi Arabia for October 12, 2026. That departure was scheduled just three days after his arrest.
He is originally from Saudi Arabia and has held lawful permanent resident status in the US since 2020.
He earned his undergraduate degree at MIT in 2016 and a master’s from UC Berkeley in 2025. Before joining OpenAI, he worked as a systems engineer at Meta. He also previously interned at Microsoft.
OpenAI hired him as a product engineer in July 2025. His tenure lasted until October 9, 2026.
If convicted, Alfozan faces a potential maximum penalty of 5 years in prison and a $250,000 fine. The charges remain allegations, and nothing has been proven in court.
Why this matters beyond one arrest
A device allegedly sat in an executive meeting room for roughly two days before a routine sweep found it. The person accused of placing it allegedly reached a space his role did not permit. Security cameras reportedly logged him near the room more than once.
OpenAI’s checks did eventually work. The routine inspection caught the first device, and the follow-up investigation reportedly led to the second. The company’s use of its own AI tools in the investigation offers a real-world example of that application in practice.
Key unanswered questions include motive and whether any recordings left the building. The research findings do not establish either.
Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.