OpenAI has arrested a 35-year-old employee, Abdulrahman Alfozan,for allegedly planting a surveillance device in a company conference room. The San Francisco-based AI giant reportedly utilized its own internal models to trace the device's origins and identify the suspect.
The toilet paper and USB port surveillance tactic
A sophisticated recording device was discovered hidden within an OpenAI office on September 30. According to a criminal complaint , the device was concealed inside toilet paper and inserted into a USB port located on the back of a television in a conference room. This method allowed the device to remain undetected until the company's security protocols flagged the anomaly.
Following the discovery, the FBI arrested Abdulrahman Alfozan, a 35-year-old Saudi national. Alfozan made his first court appearance in San Francisco on Tuesday, facing a federal charge of intercepting oral communications. The investigation revealed that Alfozan had used a productivity tool to make several inquiries regarding the company's meeting room capacities, calendars, and building addresses.
How internal AI models unmasked Abdulrahman Alfozan
OpenAI utilized its proprietary artificial intelligence models to assist in the criminal investigation into the security breach. The company reported that these internal models helped investigators understand the technical nature of the device and trace its origins back to a specific employee. This marks a singificant real-world application of AI as a defensive security tool against sophisticated internal threats.
Closed-circuit television footage provided additional evidence of Alfozan's movements within the office. While the specific conference room lacked a camera at the time the device was installed, office-wide CCTV showed Alfozan near the room on at least six occasions last month. Furthermore, a newly installed camera reportedly captured Alfozan entering the room and checking behind the television where the device had been hidden.
A mounting list of security and personnel crises at OpenAI
The arrest of Alfozan follows a period of heightened instability and scrutiny for OpenAI. The company recently faced internal friction after firing three researchers—Jasmine Wang,Tomek Korbak, and Mikita Balesni—following an investigation into the handling of sensitive information. The dismissed researchers argued that these abrupt actions could undermine a culture of safety and transparency within the firm.
This security incident occurs as OpenAI CEO Sam Altman continues to advocate for global AI safeguards before international bodies like the UN Security Council.. However, the company is simultaneously managing various internal pressures, including recent reports regarding shifts in annualized revenue and the ongoing need to protect its intellectual property from both nation-state actors and insider risks.
The unverified status of model weights and source code
Several critical questions remain regarding the ultimate objective of the alleged espionage. while OpenAI stated there are currently no indications that Alfozan inappropriately accessed or exfiltrated model weights, source code, or customer data, the full extent of the breach is still being determined by the FBI. The criminal complaint has not yet detailed what the suspect intended to do with any captured communications.
Investigators are also looking into a separate incident where Alfozan left an unattended bag in a different part of the building. According to the FBI, that bag contained his identification,a wallet, and a second recording device, suggesting a potentially broader attempt to monitor company activities.
Comments 0