OpenAI Engineer Arrested for Bugging Office, Company Says AI Caught Him

In short

The FBI arrested OpenAI engineer Abdulrahman Alfozan on Oct. 9 for allegedly planting a hidden audio recorder in an government convention room on the OpenAI’s San Francisco workplaces.
Prosecutors say a chatbot account tied to his partner requested learn how to construct, disguise, and remotely entry a steady audio recorder days earlier than the machine was put in.
OpenAI says its fashions helped establish him and that it has no indication he took mannequin weights, supply code, or buyer information.

An OpenAI engineer is in federal custody after allegedly hiding an encrypted audio recorder behind a tv in one of many firm’s government convention rooms in San Francisco. OpenAI says its personal AI fashions helped determine who put it there.

“We found a complicated recording machine that was hidden inside considered one of our workplaces,” OpenAI stated in an announcement on X. “We used our fashions to know the machine and the place it had come from, figuring out {that a} particular worker had planted it just a few days earlier.”

Myriad: Which company will have the best AI model? Click to make your prediction.

The Justice Division identified the suspect as Abdulrahman Alfozan, a 35-year-old Saudi nationwide who has labored as an engineer at OpenAI since July 2025. He was arrested on Oct. 9, charged with utilizing or making an attempt to make use of a tool to intercept oral communications, and made his preliminary look in federal courtroom on Tuesday.

“We referred the matter to the FBI, and we’re grateful to the brokers and prosecutors who moved rapidly to make an arrest,” OpenAI stated. “The FBI’s investigation is ongoing and we’ll proceed to assist their work.”

A be aware from our analysis leaders:

Final week we parted methods with Jasmine, Mikita, and Tomek after an intensive investigation discovered they violated clear insurance policies on dealing with delicate data. Our inner investigation uncovered a major breach of belief past what’s…

— OpenAI Newsroom (@OpenAINewsroom) October 9, 2026

In accordance with the felony criticism, OpenAI employees discovered the recorder on or round Sept. 30, plugged right into a USB port on the again of a TV and wrapped in bathroom paper. Its contents have been encrypted, and the corporate decided it had been put in round Sept. 28.

Safety cameras allegedly caught Alfozan close to the room on a number of dates earlier than the invention, though he had no authorization to enter it or attend its conferences. Prosecutors say these conferences recurrently concerned confidential technical and strategic data.

Earlier than the machine went in, Alfozan allegedly used an inner productiveness instrument to ask about OpenAI’s constructing addresses, assembly rooms, room capacities, calendars, and the place management and out of doors companions met.

Prosecutors additionally level to a chatbot account tied to Alfozan’s partner, which shared a number of identifiers along with his personal account. Between Sept. 24 and 30, it allegedly requested learn how to construct a steady audio recorder, add Wi-Fi connectivity, enhance stay listening, and encrypt the machine’s SD card. The criticism does not say whose chatbot it was.

“Presently, we’ve got no indications of this particular person inappropriately accessing or exfiltrating mannequin weights, supply code, or buyer information,” the corporate stated, referring to the skilled parameters that make up an AI mannequin, its underlying code, and person data. “The worker was terminated.”

BitcoinBTC · USD

$83,756−0.41%

Oct 7Oct 8Oct 10Oct 12Oct 14

$84.3k$83.1k$81.9k$80.7k

24h ExcessiveExcessive$84,074

24h LowLow$82,919

VolVol$1.1B

Market projectionsOdds by Myriad

→

“We take defending our know-how, information, and other people significantly, and as a matter of nationwide significance,” OpenAI added, saying its safety program is constructed to defend towards threats starting from insiders to cybercriminals to nation states. “This case is an instance of how AI will help handle these subtle safety threats.”

The corporate didn’t clarify in its assertion precisely what its fashions analyzed to establish him.

Federal officers are treating it as a nationwide safety case. FBI Director Kash Patel stated Alfozan allegedly tried to seize delicate conversations in regards to the firm’s “tremendous intelligence know-how,” and U.S. Lawyer Craig Missakian referred to as the prosecution a nationwide safety matter.

The arrest got here days after OpenAI’s newsroom account defended firing three security researchers—Jasmine Wang, Tomek Korbak, and Mikita Balesni—saying an investigation discovered that they had violated clear insurance policies on dealing with delicate data.

OpenAI’s safety report has drawn scrutiny since July, when the corporate admitted its personal fashions escaped a test environment and breached Hugging Face’s infrastructure to seize benchmark solutions.

If convicted, Alfozan faces as much as 5 years in jail, a $250,000 fantastic, and three years of supervised launch. A detention listening to is ready for Oct. 19 earlier than U.S. Justice of the Peace Choose Alex G. Tse.