OpenAI Engineer Arrested Over Hidden Recording Device in San Francisco Executive Conference Room

Federal prosecutors arrested an OpenAI engineer on suspicion of hiding an audio recording device in an executive conference room at the company’s San Francisco headquarters. The U.S. attorney’s office in San Francisco announced the case on Tuesday, saying agents took Abdulrahman Alfozan, 35, into custody on Friday, Oct. 9. Alfozan, a Saudi national, faces a charge of using or attempting to use a device to intercept oral communications. The case lands amid a global race to dominate artificial intelligence, where confidential research and strategy carry enormous value.

What Prosecutors Say Happened

Alfozan made his initial appearance in federal court on Tuesday and remains in custody. Prosecutors say he has worked as an engineer at OpenAI since July 2025. OpenAI personnel found the device on Sept. 30, plugged into a USB port on the back of a television in an executive conference room.

Someone had concealed the device in toilet paper, according to the criminal complaint. The company determined that the installation happened around Sept. 28, and the device’s contents were encrypted. Prosecutors say meetings in the room frequently involved confidential technical, commercial, and strategic information, including proprietary research, product plans, and government engagements.

Surveillance footage shows Alfozan near the room on multiple dates before the discovery. Prosecutors say he had no authorization to enter the room or attend meetings there. A security sweep two days after the installation uncovered the device, according to the Justice Department.

Chatbot Prompts Became Key Evidence

The complaint alleges that Alfozan used a productivity tool before the installation to ask about OpenAI’s building addresses, meeting rooms, room capacities, and calendars. He also allegedly asked about the locations of leadership and outside partner meetings. Prosecutors say those queries gave him a map of where sensitive conversations took place.

A chatbot account tied to Alfozan’s spouse, which shared identifiers with his own account, then submitted prompts between Sept. 24 and Sept. 30. The prompts covered building, installing, obscuring, and remotely accessing a recording device. Prosecutors say the queries included how to “build a continuous audio recorder” and how to “encrypt the SD card.”

The FBI’s account adds that the account also asked about “customiz[ing] wifi … connectivity” and “improv[ing] audio and live listening.” On Sept. 27 and 28, the account allegedly asked about the specific device that investigators later found, including its unique MAC address. Investigators say those references link the prompts directly to the hardware in the conference room.

A Second Device in an Unattended Bag

On Oct. 5, the complaint says, Alfozan entered the conference room and appeared to look behind the television. Surveillance footage captured the moment. That same day, he allegedly left a bag unattended in a different room.

Security personnel who opened the bag found his wallet and identification. They also found a second recording device that investigators describe as identical or similar to the first. Prosecutors say the discovery tied Alfozan to both devices and gave agents the evidence they needed to move toward an arrest.

The Justice Department says the first device was hidden in a way meant to avoid detection. The agency also says surveillance footage documented Alfozan walking around the meeting room in the days leading up to the discovery. He has not made a public comment on the indictment.

OpenAI’s Response

OpenAI said it discovered the device and referred the matter to the FBI. The company said it even used one of its own models in the process. In its statement, the company described how it handled the investigation and what it has found so far.

“We discovered a sophisticated recording device that was hidden inside one of our offices” and “referred the matter to the FBI,” OpenAI told AFP in a statement.

OpenAI also said it used one of its “models to understand the device and where it had come from.” The company identified “that a specific employee had planted it a few days earlier.” It then moved to reassure customers and partners about the security of its core assets.

“At this time, we have no indications of this individual inappropriately accessing or exfiltrating model weights, source code, or customer data. The employee was terminated,” OpenAI said.

The company added that it was “grateful to the agents and prosecutors who moved quickly to make an arrest.” OpenAI said the investigation is continuing.

Charges and Potential Penalties

If a court convicts Alfozan, he faces up to five years in prison and fines of up to $250,000. The charge centers on the interception of oral communications, not on any confirmed theft of company data. Prosecutors have not said what, if anything, the first device recorded before security found it.

The case highlights how physical security has become as important as digital defenses for leading AI companies. Executive conference rooms at these firms host discussions about proprietary research, product plans, and government work. Investigators say this episode combined old-fashioned hardware with modern chatbot tools, and the chatbot prompts now form a central part of the evidence.

For now, the allegations remain unproven, and Alfozan has the right to contest them in court. Federal agents and prosecutors continue to investigate, and OpenAI says it is working with them. The company says it takes the protection of its technology and data seriously.

Image: AI-generated