An OpenAI engineer has been arrested on a federal charge of using a device to intercept oral communications, after a hidden recorder was allegedly found during a security sweep of an executive conference room. Abdulrahman Alfozan, 35, a Saudi national who joined OpenAI in July 2025, was arrested on Friday and made an initial federal-court appearance on Tuesday, the San Francisco Standard reported. He remained in custody ahead of a detention hearing on October 19 before U.S. Magistrate Judge Alex G. Tse. If convicted, he faces up to five years in prison, a $250,000 fine and three years of supervised release.
Found in a sweep, hidden in toilet paper
According to the federal complaint, OpenAI personnel discovered the device on September 30 during a technical surveillance countermeasures sweep. It was plugged into a USB port on the back of a TV in an executive conference room and concealed in toilet paper. Prosecutors allege it was installed around September 28 and that its contents were encrypted. The meetings held in that room allegedly involved confidential technical, commercial and strategic information — proprietary research, product plans and government engagements.
The complaint further alleges that surveillance showed Alfozan near the room on multiple dates without authorisation, and that before the installation he had queried an internal productivity tool about building addresses, meeting rooms, capacities, calendars and where leadership and partner meetings take place.
The chatbot trail
Most strikingly for an AI company, part of the alleged evidence is a chatbot history. A chatbot account tied to Alfozan’s spouse, sharing identifiers with his own account, allegedly prompted between September 24 and 30 about building, installing, obscuring and remotely accessing a recorder and encrypting an SD card — and on September 27–28 about the specific device model and its MAC address.
On October 5, prosecutors say, Alfozan re-entered the room and looked behind the TV, and left a bag unattended in another room; security found his wallet and ID in it, along with a second, similar recording device. OpenAI has said it used its own AI models to identify and analyse the device, referred the case to the FBI, progressively restricted his access from September 30, placed him on leave, and terminated him on Friday in coordination with the FBI. The company says it has no indication he accessed model weights, source code or customer data.
“A matter of national security”
U.S. Attorney Craig Missakian called the case “a matter of national security”, and FBI leadership framed the protection of U.S. AI innovation as a top counterintelligence priority. The prosecution is being handled by the National Security, Cyber & Special Prosecutions Section of the U.S. Attorney’s Office for the Northern District of California — a signal of how seriously the government treats alleged insider surveillance at a frontier AI lab. A LinkedIn profile matching Alfozan describes an engineering lead on an AI agents platform, previously a founding engineer at Nash and a senior systems engineer at Meta, with a master’s in data science from UC Berkeley. His family, at his Millbrae home, declined to speak to reporters.
The physical layer is the weak layer
Frontier labs spend fortunes on cybersecurity and model-weight protection. This case is a reminder that the most sensitive room in the building can still be compromised through the back of a television — and that the same AI tools the industry builds can end up as evidence, with a chatbot history allegedly documenting the plot. It sits alongside the other always-on listening questions we have covered, from the hidden risks of AI assistants to the compute build-out behind NVIDIA and Cisco’s optical modules for AI data centres.
Alfozan is presumed innocent, and the allegations will be tested in court. But every AI lab’s security team will be re-checking its own conference rooms this week — starting with the TVs.
