OpenAI has been training employees to spot espionage attempts, from romance "honey pots" to strangers asking about their work, The Wall Street Journal reported on October 14, citing unnamed sources. Staff are told to avoid company logos in public, hide their screens and wipe whiteboards after meetings.
The report came one day after the US Justice Department announced the arrest of an OpenAI engineer accused of hiding an audio recording device in an executive conference room. The Journal reports that OpenAI was already warning staff about spies before the arrest.
What OpenAI tells employees
According to the Journal, as reported by the New York Post:
- Honey pots. Trainings warn that foreign agents may use sex or romance to get close to a target and steal secrets.
- Other methods. Security staff list bribes and eavesdropping on employees in public places such as cafes.
- A real approach. A stranger made small talk with an OpenAI employee at a San Francisco playground, then turned to their work. The Journal reports the stranger was a spy.
- Everyday rules. Employees are told to avoid wearing clothes with company logos in public, use screen protectors to hide phones and laptops in public, and erase whiteboards after meetings.
The Journal also reports that Anthropic and Google have given their employees similar warnings.
The arrest
The Justice Department announced the case on October 13, 2026:
- The engineer was arrested on October 9 and is charged with using or trying to use a device to intercept oral communications.
- On or about September 30, OpenAI staff found the device in a USB port on the back of a television in an executive conference room in San Francisco.
- The charge carries a maximum penalty of 5 years in prison and a $250,000 fine.
A charge is an accusation. He has not been convicted.
Sources
Did this land? Leave a reaction.
Discussion
0 commentsMembers can join the discussion
Become a member to comment and reply to other members.
Become a MemberNo comments yet
Be the first to start the discussion.