A Saudi national and OpenAI employee, Abdulrahman Alfozan, has been arrested in San Francisco on charges that he planted a covert audio recording device in an executive conference room to secretly capture sensitive conversations. He is accused of using or attempting to use an electronic device to intercept oral communications.
The arrest followed OpenAI’s discovery of a surreptitious recording device inserted into a USB port on the back of a TV in the room, believed to have been installed around 28 September 2026 and found around 30 September. CCTV footage reportedly shows Alfozan in proximity to the conference room on several dates prior to the discovery, despite having no authority to attend those discussions. He had been employed at OpenAI since July 2025 as an engineering lead and was terminated after the incident.
OpenAI publicly stated that the device was sophisticated, containing encryption, audio monitoring capabilities, and a data-transfer mechanism designed to evade monitoring. The company said there is no evidence so far that model weights, source code, or customer data were accessed or exfiltrated. The FBI and DoJ affidavit quote statements emphasising the significance of protecting sensitive AI innovations.
If convicted, Alfozan faces a potential maximum penalty of five years in prison and a $250,000 fine, followed by three years of supervised release. The case follows recent OpenAI actions, including the firing of three safety researchers earlier in October after internal policy breaches.