OpenAI has warned employees that competitors and state-sponsored actors could try to obtain company information, the Wall Street Journal reported on 15 October. Its training gives staff examples of how an approach might happen, including one at a playground.
In that scenario, a stranger repeatedly turned up with a child at a San Francisco playground where an OpenAI employee was with their child. The stranger struck up conversations and asked about the employee’s work. The Journal described the stranger as a spy.
Company leaders also warned staff about offers of money, people listening in cafés and attempts to seduce employees, the report said. The warnings preceded federal authorities’ arrest of an employee over a listening device. AI Affairs previously reported the arrest and charge of an OpenAI engineer over a hidden office recorder.
Employees at Anthropic and Google have received similar espionage warnings, according to the Wall Street Journal.
