A hidden recording device in the room where executive meetings are held and company secrets are discussed: AI giant OpenAI announced it faced exactly such a security breach at its San Francisco headquarters.

The device was connected to the USB port behind the room's television and camouflaged with toilet paper. The room was where highly confidential matters — from product roadmaps to research efforts, even talks with the government — were on the table.

ope

Company used its own models to identify the culprit

According to the official statement, OpenAI deployed its own AI models to examine the seized device. The company said this analysis allowed it to determine who planted the device within days. It stressed, however, that initial checks found no trace of unauthorized access to or exfiltration of model weights, source code, or customer data.

According to information in the investigation file, the device was placed in the room around September 28 and discovered on September 30. Suspicion fell on 35-year-old Saudi national Abdulrahman Alfozan, an engineer at the company since July 2025. Alfozan had no authorization to enter or use the executive room; security cameras, however, showed him repeatedly approaching the area near where the device was placed.

ope2

A second device turned up in a bag

Alfozan was fired after the incident, and after the file was referred to the FBI, he was taken into custody, appearing before a judge at his first hearing on Tuesday. If convicted, he faces up to 5 years in prison, a $250,000 fine, and 3 years of supervised release.

On October 5, security staff reached a bag the suspect had left behind inside the company. Along with his ID and wallet, it contained a second listening device. According to the allegations, searches such as "how to make a continuous audio recorder" and "how to encrypt an SD card" were made on a chatbot account belonging to the suspect's wife. No explanation has been given for why Alfozan carried out the act.