'Honey Pot' Spy: OpenAI Warns Engineers of New Tactics as Espionage Threats Escalate

OpenAI is warning its engineers that foreign spies could use "honey pot" schemes, bribery, eavesdropping and seemingly casual encounters to steal AI secrets, underscoring the growing espionage risks facing the artificial intelligence industry. The warnings come as OpenAI deals with an alleged insider surveillance case involving engineer Abdulrahman Alfozan, who was arrested after a listening device was discovered inside an executive conference room. According to a report by The Wall Street Journal, OpenAI has been training employees to recognize tactics designed to exploit personal relationships and everyday situations. In one reported incident, a stranger approached an OpenAI employee who was with their child at a San Francisco playground, initially making small talk before steering the conversation toward the employee's work at the AI company. The stranger was allegedly a spy.