© 01net
A forgotten bag containing an ID card, incriminating online searches, and AI models let loose on the investigation: the criminal case targeting an OpenAI employee arrested by the FBI borders on an espionage parody.
At the end of September, OpenAI’s security teams discovered an audio recorder inserted into a USB port behind the television in a meeting room reserved for senior management in San Francisco. The device, which the company deemed “sophisticated”, was hidden in toilet paper and had been installed approximately two days earlier.The internal investigation lasted about ten days before yielding a name: that of a company employee.
Toilet paper and a forgotten bag
The suspect’s online searches, unearthed by investigators, focused on the manufacture, customization, installation, concealment and remote access of a continuous audio recorder. A photograph added to the case file also shows him handling a recorder on the premises.
A bag left unattended in another room in the building contained the rest of the evidence: a second recorder, stored alongside Abdulrahman Alfozan’s wallet and ID card. The Saudi national is said to have joined the product team in July 2025. According to the evidence gathered by the company, more devices were expected to follow. Arrested on October 9, the man was dismissed shortly afterward and charged with using a device designed to intercept communications. No mobile phone has been made public, and no client has been identified for the time being.
The in-house models deployed in the investigation
To trace the lead, OpenAI used its own artificial intelligence models, tasked with “understanding the device and where it came from.” The company then referred the matter to the FBI, whose investigation is ongoing.
“At this stage, we have no evidence of inappropriate access or exfiltration of model weights, source code, or customer data by this individual,” the company assured.
Already in early October, three employees, including two security researchers, had been fired for having violated the internal rules for handling sensitive information. In 2023, a hacker’s intrusion into its internal forums was never reported to the FBI. This time, OpenAI made the call itself. For AI laboratories, custodians of coveted secrets, the threat now comes knocking even in meeting rooms.
👉🏻 Follow tech news in real time: add 01net to your sources on Google, and subscribe to our WhatsApp channel.
Source: Gizmodo
- Google Play Store
- 4K Video Downloader Plus
- VLC Media Player
- Adobe Reader
- uTorrent
- WinRAR
- Xender
- Google Chrome
- CCleaner
- Mozilla Firefox
- Avast Free
- ChatGPT
- Microsoft 365
- Opera One
- Avast One










- Google Play Store
- 4K Video Downloader Plus
- VLC Media Player
- Adobe Reader
- uTorrent
- WinRAR
- Xender
- Google Chrome
- CCleaner
- Mozilla Firefox
- Avast Free
- ChatGPT
- Microsoft 365
- Opera One
- Avast One






