In a bizarre intersection of artificial intelligence and corporate espionage, a former OpenAI engineer has been arrested for allegedly using the company’s own technology to orchestrate a surveillance plot within its headquarters. Federal prosecutors claim that Abdulrahman Alfozan, a 35 year old Saudi national and U.S. resident, utilized his wife’s ChatGPT account to generate detailed instructions for building a covert audio recording device. The prompts acted as a digital blueprint, guiding him through the creation of a continuous recorder equipped with live listening capabilities and an encrypted memory card designed to hide its tracks.
The scheme reportedly began with careful reconnaissance. Court documents reveal that Alfozan first used an internal OpenAI productivity tool to identify high stakes meeting rooms capable of hosting large groups and frequenting visits from external partners. His goal was apparently to target conversations involving highly confidential technical research and strategic business developments. Once the ideal location was identified, he allegedly planted a thumb sized recording board, wrapped in toilet paper, behind a television in an executive conference room where only top leadership had authorized access.
Company security eventually discovered the device during a routine sweep, sparking an investigation involving the FBI. While investigators were unable to decrypt any files on the recovered memory card, they built their case through extensive CCTV footage showing Alfozan lingering near the room throughout September. Suspicion intensified when security searched a bag Alfozan had left in a printer room and found a second identical recording board alongside his personal belongings. Despite later claiming several items were stolen from the bag, he notably failed to mention the missing spy hardware.
Federal agents moved quickly to apprehend Alfozan on October 9, just three days before he and his family were scheduled to depart on one way flights back to Saudi Arabia via Istanbul. He now faces charges under the federal Wiretap Act, which carries a potential sentence of five years in prison. Although OpenAI stated there is currently no evidence that sensitive source code or customer data was compromised, the involvement of the Justice Department’s National Security Division suggests that authorities are treating the breach with significant gravity regarding American intellectual property protections.