OpenAI warns employees that spies could use sex and bribes to steal its AI secrets

OpenAI is warning its employees that foreign spies could use romantic advances, sexual relationships and financial incentives to gain access to confidential company information, according to a report by The Wall Street Journal. The warnings come as concerns over the security of sensitive AI research grow, following the discovery of a hidden recording device inside an executive conference room at the company's San Francisco headquarters.The AI company has reportedly been educating its employees about so-called “honey pot” schemes, in which intelligence operatives attempt to get close to their targets through romantic or sexual advances before seeking access to sensitive information.One incident cited in the report involved an OpenAI employee spending time with their child at a playground in San Francisco. A stranger approached the employee and started a casual conversation, but eventually began asking questions about their work at the AI company. The person was allegedly a spy attempting to gather information.OpenAI warns of honey pot schemes and other spying tacticsDuring security training, OpenAI employees are reportedly told that foreign intelligence agents could use sexual attraction to build relationships with employees and extract confidential information. The company has also used an example involving a US diplomat stationed in Russia to explain how such tactics work. According to the report, foreign agents allegedly sent a blonde woman carrying Champagne to the diplomat's hotel room. After he refused to let her in, they reportedly tried again by sending a redheaded woman and later a young man. Sexual advances are not the only threat employees have been warned about. According to the report, OpenAI's security experts have also discussed the possibility of spies offering bribes or trying to overhear conversations in public places, including cafes.Employees have reportedly been advised to take precautions when working outside the office. These include using privacy screen protectors on phones and computers, clearing whiteboards after meetings and avoiding clothing with company logos in public. OpenAI is not the only company taking such precautions. Other major AI developers, including Google and Anthropic, have reportedly issued similar warnings to their employees as competition in the industry intensifies.Hidden recording device raises security concernsThe warnings come days after US authorities arrested Abdulrahman Alfozan, a 35-year-old Saudi national and former OpenAI engineer, over allegations that he planted a recording device inside the company's headquarters. The discovery prompted a federal investigation into how the recording device had been placed inside OpenAI's office. US authorities have accused former employee Abdulrahman Alfozan of attempting to capture private conversations. Investigators say the device was found in an executive meeting room, concealed in a way that made it difficult to spot.Alfozan, 35, was arrested last Friday. The charge carries a maximum penalty of five years in prison and a $250,000 fine. Following the arrest, FBI Director Kash Patel said protecting confidential information related to AI development was essential to the country's economic strength and national security. He added that counterintelligence efforts remained a major priority for the agency.- Ends