Federal agents arrested Abdulrahman Alfozan on October 9, accusing the 35-year-old OpenAI engineer of intercepting oral communications at the company’s San Francisco offices. Prosecutors say he hid a recorder, wrapped in toilet paper, behind a television in a conference room; they also cited alleged chatbot prompts from his wife about building and concealing an audio device.¹

Alfozan was later indicted by a grand jury on October 27 on one count of interception of oral communications. The government’s account centers on an alleged September 28 placement of a device inside a television in an executive conference room — an accusation that carries a potential maximum five-year prison sentence if proven.²

At a Friday hearing in San Francisco federal court, Alfozan entered a not-guilty plea through attorney David Rizk, who waived a reading of the indictment. The plea puts the prosecution’s allegations squarely in dispute; no evidence has yet been tested at trial.

For now, however, the immediate fight is not over guilt but custody. At an earlier hearing, a judge found Alfozan to be a flight risk and denied bond. He was returned to jail after Friday’s brief appearance.

The detention contrasts with a personal history his prior lawyer has emphasized: Alfozan has lived in the United States since college, held a green card, and was reportedly weeks from a naturalization interview when he was arrested. Before OpenAI, he worked at Facebook, the startup Nash and Saudi venture firm Raed Ventures.

The case lands amid sharper US concern over the protection of advanced AI technology, while OpenAI has also held discussions with Saudi Arabia’s sovereign wealth fund about a possible investment. Alfozan’s next court hearing is scheduled for November 17.