Federal prosecutors have charged Abdulrahman Alfozan, a Saudi national who has worked as an engineer at OpenAI since July 2025, accusing him of attempting to intercept private conversations by placing a hidden recording device in an executive conference room.
According to company reports, OpenAI discovered a “surreptitious audio recording device” around Sept. 30 that had been plugged into a USB port at the back of a television in an executive meeting room. OpenAI said on X that it employed its own AI models to help identify the device and trace its origin.
Allegations in the investigation
An FBI affidavit filed in the case states investigators believe Alfozan used AI tools both to research the company’s meeting spaces and to develop the recording equipment. The complaint cites queries made to a so-called “productivity agent,” including: “What meeting was this presented at and when was it record?,” “List meeting rooms in (company) that are 6+,” and “can you list today's meeting in this room.”
The documents also say an AI chatbot account linked to his spouse asked the model to build a “continuous audio recorder,” add “browser access and controls,” and encrypt the memory card where recordings would be stored. The complaint does not say whether any sensitive conversations were successfully recorded.
Why this matters
The case highlights a growing concern for companies developing frontier AI: employees with access plus powerful AI tools can create new avenues for insider threats. The same models firms race to build and deploy can be misused to plan and execute wrongdoing, including alleged corporate espionage.
This incident is the latest example of AI being implicated in criminal cases. Prosecutors previously identified a Nebraska man’s use of ChatGPT in the lead-up to a bank robbery; that individual was sentenced in August to more than ten years in prison. Authorities also said last year that someone used ChatGPT to help plan an explosion of a Tesla Cybertruck near the Trump International Hotel in Las Vegas.
In a related strand of cases, the Justice Department in 2024 indicted former Google software engineer Linwei Ding for stealing AI technology secrets and providing them to two Chinese companies; a federal jury this year found Ding guilty on several counts of economic espionage and theft of trade secrets.
Risks and implications
Authorities note AI tools can play a dual role: they can aid companies’ internal investigations—as OpenAI reported doing in analyzing the discovered device—but they can also lower the technical barrier for would-be perpetrators seeking to build spying hardware or plan illicit acts. That creates new challenges for corporate security policies, access controls, and legal accountability.
Further developments in the case will clarify whether any confidential conversations were recorded and how central AI tools were to the alleged plot. The episode underlines that generative AI raises not only technological questions but also legal and security concerns for organizations operating sensitive internal environments.
