An OpenAI employee has been arrested following the discovery of a sophisticated recording device in one of the company's offices. OpenAI stated it referred the matter to the FBI.
In a statement on Tuesday, the San Francisco-based AI company confirmed it used one of its models to understand the device and its origin. OpenAI identified that a specific employee had planted the device a few days prior and subsequently terminated their employment.
The company confirmed the employee's arrest and stated that there are currently no indications the individual inappropriately accessed or exfiltrated model weights, source code, or customer data. OpenAI did not disclose details on how its models were used to identify the employee or the specific office where the device was found.
OpenAI expressed gratitude to the agents and prosecutors involved, noting that the investigation is ongoing. The company emphasised its commitment to protecting its technology, data, and personnel, describing it as a matter of national importance.
This incident follows other privacy-related concerns for OpenAI, including a hacking incident involving Australia's university healthcare system in June and an investigative subpoena from the California attorney general earlier this month regarding rogue hacking. Last week, OpenAI also terminated three safety researchers for allegedly sharing sensitive information.