B2 Reading: A Bug in the Boardroom, Spying at OpenAI?
Read the article about the hidden listening device at OpenAI and fill in the gaps.
Word Bank
Tap a word, then tap a gap. On a computer you can also drag it.
claim
suspected
sophisticated
innocent
arrested
detected
charged
identify
spying
confidential
surveillance
security
It sounds like the plot of a spy film. On 30 September 2026, staff at OpenAI's San Francisco headquarters noticed a small device plugged into a USB port on the back of a television in an executive conference room. Wrapped, according to court documents, in toilet paper, it turned out to be a listening device, with encrypted storage and a system for sending recordings out of the building without being .
The location was no accident. Prosecutors say the room is regularly used for leadership meetings and for briefings with government officials and business partners, and that OpenAI executives had met government officials there in the days around the time the device was installed.
What happened next added an unusual twist. OpenAI says it used its own AI models to help investigators examine the device and the person of planting it: a 35-year-old engineer who had joined the company in July 2025. He was fired and by the FBI on 9 October, just days before he and his family were booked on a one-way flight out of the country. He has been with intercepting private conversations.
According to prosecutors, the engineer had used a workplace tool to research meeting rooms and executives' calendars, and an AI chatbot to get instructions on building and hiding the device. They also that he had been in contact with a member of the Saudi royal family, which is why the US government is treating the case as a matter of national . None of these claims has been tested in court yet, and the law treats him as until he is proven guilty.
The case shows how valuable the secrets of leading AI companies have become. Their research, business plans and partnerships with governments are now seen as extremely valuable, and experts have warned for years that AI labs are attractive targets for corporate and government . Ironically, in this case, AI may also have helped to catch the alleged spy.
For other companies, the story raises uncomfortable questions. How far should employers go to protect their secrets? And how much of their own staff are workers willing to accept?
The location was no accident. Prosecutors say the room is regularly used for leadership meetings and for briefings with government officials and business partners, and that OpenAI executives had met government officials there in the days around the time the device was installed.
What happened next added an unusual twist. OpenAI says it used its own AI models to help investigators examine the device and the person of planting it: a 35-year-old engineer who had joined the company in July 2025. He was fired and by the FBI on 9 October, just days before he and his family were booked on a one-way flight out of the country. He has been with intercepting private conversations.
According to prosecutors, the engineer had used a workplace tool to research meeting rooms and executives' calendars, and an AI chatbot to get instructions on building and hiding the device. They also that he had been in contact with a member of the Saudi royal family, which is why the US government is treating the case as a matter of national . None of these claims has been tested in court yet, and the law treats him as until he is proven guilty.
The case shows how valuable the secrets of leading AI companies have become. Their research, business plans and partnerships with governments are now seen as extremely valuable, and experts have warned for years that AI labs are attractive targets for corporate and government . Ironically, in this case, AI may also have helped to catch the alleged spy.
For other companies, the story raises uncomfortable questions. How far should employers go to protect their secrets? And how much of their own staff are workers willing to accept?