A Saudi national employed as an engineer at OpenAI has been arrested for allegedly placing a hidden recording device in a conference room at the company’s San Francisco office, federal officials announced on Tuesday.
Abdulrahman Alfozan, 35, was taken into custody on October 9 and faces charges for using or attempting to use a device to intercept oral communications, as stated by the Department of Justice.
FBI Director Kash Patel mentioned on social media that Alfozan was arrested for allegedly hiding a recording device in an executive meeting space to secretly capture sensitive discussions about the company’s advanced technology. He emphasized the importance of protecting American innovation in artificial intelligence, which the FBI considers a top counterintelligence priority.
OpenAI discovered the encrypted recording device on September 30, connected to a television in a conference room designated for “highly confidential” discussions regarding proprietary research, product plans, and business strategies. The device had reportedly been hidden in toilet paper and attached two days prior.
Surveillance footage revealed that Alfozan had accessed the conference room multiple times without proper authorization before the device was found. In addition, he had conducted searches related to OpenAI’s facilities, meeting rooms, and executive schedules.
A chatbot account linked to Alfozan’s spouse, which had shared identifiers with his account, allegedly submitted queries about building and concealing recording devices with features like continuous audio recording and live listening capabilities.
On October 5, the security footage captured Alfozan entering the conference room and seemingly inspecting the area behind the television where the device had been concealed. That same day, a second recording device was discovered in a bag that Alfozan had left unattended in another room, alongside his wallet and ID.
OpenAI indicated that they utilized their technology to determine the origin of the device, identifying Alfozan as the individual who had planted it days earlier. The company reported no evidence that he improperly accessed or extracted sensitive data and subsequently terminated his employment.
Amy Kremer, chairwoman of Humans First, expressed that the situation underscores the risks of insider threats in organizations working on cutting-edge AI technologies. She pointed out the necessity for robust security measures in these companies to ensure national security, calling for thorough investigations into the allegations made against Alfozan.
Alfozan made his initial court appearance on Tuesday, and if found guilty, he could face up to five years in prison, alongside a $250,000 fine and a period of supervised release. OpenAI has not yet provided additional comments on the situation.
