Saudi national charged with planting a recording device at OpenAI
The Justice Department said Abdulrahman Alfozan, an OpenAI engineer, hid a device in an executive conference room and had a second one in his bag when caught.
- Security & misuse
- Labs & people
- Notable
The US Department of Justice announced charges against Abdulrahman Alfozan, a 35-year-old Saudi national who had worked as an engineer at OpenAI since July 2025, for using a device to intercept oral communications. He was arrested on 9 October and made his first federal court appearance in San Francisco on 13 October.
According to the unsealed complaint, OpenAI staff found a recording device hidden in toilet paper inside a USB port on a television in an executive conference room on 30 September, installed roughly two days earlier; its contents were encrypted. CCTV footage showed Alfozan near the room on multiple occasions, though prosecutors said he had no authority to attend meetings held there, which the complaint describes as covering confidential research, product plans, business development and dealings with government partners. The complaint also alleges that a chatbot account linked to his spouse submitted prompts in late September about building, customising and concealing an audio recorder, and about the specific device later found, and that on 5 October — the same day he was seen looking behind the television — he left a bag containing a second, similar recording device unattended elsewhere in an OpenAI building, where security staff found and opened it.
Officials framed the case in national-security terms rather than as routine workplace misconduct. FBI Director Kash Patel said the bureau views “protecting American SI [super intelligence] innovation” as one of its highest counterintelligence priorities, and US Attorney Craig Missakian for the Northern District of California called it “a matter of national security.” Director of National Intelligence Jay Clayton linked the arrest to the administration’s Super Intelligence Force. The prosecution is brought in support of National Security Presidential Memorandum 11, a directive aimed at maintaining US advantage in frontier AI. If convicted, Alfozan faces up to five years in prison and a $250,000 fine.
The Justice Department’s statement did not say what, if anything, had been recovered from the devices, who Alfozan might have been working for, or whether he remains employed by OpenAI. Writing in Transformer, James Ball argued the case exposed a broader problem: that the physical and personnel security culture at fast-growing AI labs has not caught up with the sensitivity of what they now hold, and that “startup culture can’t guarantee the security OpenAI, Anthropic and the rest need.”
In the commentary
What people were saying around this time — external links, from the record's commentary rail.
- 13 October 2026 · Shakeel Hashim · Transformer‘Artificial’ is not the AI film we deserve
- 14 October 2026 · Richard MacManus · Latent SpaceCrossing the Threshold: How OpenAI and Anthropic Approach Autonomy and Safety in Financial Agents
- 12 October 2026 · Zvi Mowshowitz · Don't Worry About the VaseOpenAI Fires Three Safety Researchers