Major artificial intelligence companies are grappling with aggressive espionage campaigns targeting their top-secret intellectual property and model architecture. Training scenarios used by firms like OpenAI highlight sophisticated tactics, including foreign operatives attempting to befriend employees at local spots like playgrounds. The heightened threat environment extends beyond social engineering, encompassing insider threats and attempts to plant listening devices in offices. Security concerns intensified following federal law enforcement actions, including the arrest of an OpenAI employee for planting a clandestine recording device. In response, Silicon Valley tech leaders are ramping up internal vigilance, tightening corporate security measures, and bolstering defenses against state-sponsored corporate

Full report : Cyber espionage and insider threats in the age of artificial intelligence and autonomous agents.