DOJ charges OpenAI engineer with secretly recording conversations

The Justice Department charged an OpenAI engineer with secretly recording conversations inside the company’s San Francisco headquarters, unsealing a criminal complaint against a Saudi national who allegedly hid a listening device inside a roll of toilet paper near an executive conference room television. The case, filed in the Northern District of California, is tied to National Security Presidential Memorandum 11, a directive aimed at protecting U.S. “super intelligence” technology.

• Abdulrahman Alfozan, 35, arrested Oct. 9, made his initial federal court appearance today
• Second recording device found in his unattended bag inside an OpenAI building on Oct. 5
• Watch for a detention or bail hearing and the government’s first substantive evidentiary filing

• 5 yrs maximum prison term if convicted on the single charged count
• $250K maximum fine listed in the complaint alongside the prison term
• 3 yrs supervised release following any custodial sentence
• Jul ’25 month Alfozan began work as an OpenAI engineer, per the complaint

According to the Justice Department release, Alfozan has been charged with using or endeavoring to use an electronic device to intercept oral communications, a single count tied to a device OpenAI personnel discovered on Sept. 30 inside a USB port on the back of a television in an executive conference room. The complaint says OpenAI determined the device was installed around Sept. 28, had been concealed in toilet paper, and stored its contents in encrypted form. Closed-circuit footage reviewed by investigators allegedly placed Alfozan near the conference room on multiple occasions before the discovery, despite what the complaint describes as his lack of authority “to access or to participate in meetings occurring in the executive conference room.”

A chatbot trail tied to a spouse’s account

The complaint’s most granular allegations involve digital research the government says preceded the installation. It alleges Alfozan used a “productivity tool” to query OpenAI’s building addresses, meeting-room capacities and calendars, and “leadership and exec” locations.

A separate chatbot account the government links to Alfozan’s spouse, which it says shared multiple identifiers with his own account, allegedly submitted prompts between Sept. 24 and Sept. 30 covering how to “build a continuous audio recorder,” customize Wi-Fi connectivity, and “encrypt the SD card.” The same account, according to the complaint, later submitted prompts referencing the recovered device’s custom header and unique MAC address between Sept. 27 and Sept. 28, days before the hardware was found.

A second device surfaces in an unattended bag

The complaint alleges that on Oct. 5, Alfozan entered the executive conference room and appeared to look behind the same television where the first device had been inserted. That same day, he allegedly left a bag unattended in a different OpenAI room.

OpenAI security personnel opened the bag and found, alongside Alfozan’s wallet and identification, a second device similar to the first, the complaint says. The government has not disclosed what, if anything, either device recorded or transmitted before discovery, nor whether any captured audio has been reviewed as evidence.

“Protecting American SI innovation is essential to safeguarding our nation’s economic strength and national security, and the FBI views that mission as one of our highest counterintelligence priorities.”

Kash Patel, FBI Director

What the complaint leaves open

The department frames the case in explicitly geopolitical terms, invoking NSPM-11 and what DNI Jay Clayton called “the global super intelligence race,” but the complaint as described does not allege that Alfozan acted on behalf of any government or outside entity. Alfozan is a charged defendant, not a convicted one, and the allegations remain to be tested in court. The release does not say whether he has entered a plea, retained counsel, or been granted bail, and it does not specify what, if any, information the devices actually captured before OpenAI found them.

What changes for AI labs and their insurers

For OpenAI and peers racing on frontier model development, the case formalizes what until now has been an informal worry: physical insider threats inside buildings that already run extensive digital security. The complaint’s reliance on chatbot prompt logs as evidence, tied to an account allegedly linked to a spouse rather than the defendant directly, shows investigators treating AI query history as a forensic data source in the same way email or browser logs have been used.

That has practical implications for enterprise AI governance teams and corporate counsel at labs handling similarly sensitive model work. Expect tighter controls on executive-floor access, more aggressive conference-room sweeps, and renewed scrutiny from boards over insider-threat programs now that a federal prosecution has put a dollar-and-years figure on the risk, even if the maximum penalty here, five years and a $250,000 fine, is modest next to the value of the intellectual property the complaint says was at stake.

The BlockWest read. This case turns AI labs’ own chatbot logs into evidence against insiders, and it puts a counterintelligence label on ordinary corporate espionage risk inside frontier AI firms. Allocators backing OpenAI-adjacent companies should expect security diligence questions to now include physical access controls and NSPM-11 compliance posture, not just model safety reviews.

The next concrete step is Alfozan’s detention or bail hearing in the Northern District of California, where prosecutors Nicolas Parker and Kate Tagert will need to lay out what, if anything, the recording devices actually transmitted.

BlockWest is a news publication. Nothing here is investment advice. Read our disclaimer and editorial policy.