OpenAI Engineer Charged Over Alleged Office Surveillance Scheme

The Discovery and the Target

OpenAI security personnel discovered the bug on September 30 during a countersurveillance sweep. The device was concealed in a USB port on the back of a television inside a conference room that can only be reserved by executives and executive assistants. The room is regularly used for leadership meetings, technical and product sessions, and briefings involving confidential information concerning research, business strategy, and government relationships.

Authorities identified the hardware as a Seeed XIAO ESP32-S3 Sense, a thumb-sized development board featuring a digital microphone, detachable camera sensor, and microSD storage. While OpenAI's security team attempted to export the data on the microSD card, they could not review it because the storage was encrypted.

Following the discovery, OpenAI notified the FBI. Investigators connected the device to Alfozan after reviewing surveillance footage, internal chat logs, and an unattended bag left in a printer room on October 5. Inside that bag, security found a second encrypted recording device alongside personal items including Alfozan's wallet. OpenAI security removed the second device before Alfozan later reported the bag missing and recovered it from the lost and found.

Company Response and Background

OpenAI stated on X that it has found no indications that the individual inappropriately accessed or exfiltrated model weights, source code, or customer data. The company emphasized that it takes the protection of its technology, data, and people seriously.

Prior to his employment at OpenAI, Alfozan reportedly worked at Meta and a cargo logistics startup, and held a part-time position as a venture partner at RAED Ventures in Saudi Arabia. Neither Meta nor RAED Ventures responded to requests for comment.

What it means for developers

Incidents involving insider threats and corporate espionage highlight the rigorous security measures tech companies implement around proprietary research and internal tools. For developers building applications, maintaining robust security protocols around data privacy and access controls remains a top priority. When experimenting with various models and integrating capabilities into projects, developers can try top AI models cheaply through one API at https://apixoai.online without needing multiple vendor accounts.

The case remains ongoing in US district court, with prosecutors continuing to evaluate the evidence gathered from the investigation.

Source: An OpenAI Employee Allegedly Used a Chatbot to Help Plant a Recording Device at the Office — Wired AI. Written by the Apixo team from that report.