An OpenAI employee allegedly used AI to spy on his own company — then AI helped catch him

(Image credit: Shutterstock) You really can't make this stuff up. An OpenAI engineer allegedly used AI tools to help plan a secret surveillance operation against his own employer. Now he's facing federal charges, and OpenAI says it used its own AI models to help investigate the device.According to a criminal complaint released by the Department of Justice, 35-year-old Abdulrahman Alfozan allegedly planted a hidden audio recorder inside an executive conference room at OpenAI's San Francisco offices.The device was plugged into the USB port behind a television, concealed in toilet paper and equipped with encrypted storage.The chatbot conversations that raised questionsWe discovered a sophisticated recording device that was hidden inside one of our offices. We used our models to understand the device and where it had come from, identifying that a specific employee had planted it a few days earlier. We referred the matter to the FBI, and we’re grateful to the agents and prosecutors who moved quickly to make an arrest. The FBI’s investigation is ongoing and we will continue to support their work. At this time, we have no indications of this individual inappropriately accessing or exfiltrating model weights, source code, or customer data. The employee was terminated. We take protecting our technology, data, and people seriously, and as a matter of national importance. Our security program is designed to protect against a wide range of threat actors, from insider risks to cybercriminals to nation states, and this case is an example of how AI can help address these sophisticated security threats.October 13, 2026Arguably the most interesting part of the story isn't even the recording device; it's the chatbot conversations investigators say preceded its discovery.Federal prosecutors allege that an AI chatbot account associated with Alfozan's spouse was used to research how to build, conceal and remotely access an audio recording device. The account also submitted questions referencing the specific device later discovered at OpenAI.Meanwhile, the company says its own AI models helped security personnel understand the suspicious hardware.The result is an unusual case in which AI allegedly helped someone prepare a surveillance operation while also helping the company investigate it.Sign up to the Tom's AI Guide weekly newsletter summing up all the biggest AI news you need to know. Plus, analysis from our AI editors and tips on how to use the latest AI tools!The alleged operation began taking shape in September (Image credit: Getty Images)According to prosecutors, the alleged operation began taking shape in September. Before the device appeared in the conference room, Alfozan allegedly used a productivity tool to search for information about OpenAI's office buildings, meeting rooms, calendars and executive meeting locations.Between September 24 and September 30, a chatbot account associated with his spouse submitted a series of technical questions about recording equipment.The prompts covered continuous audio recording, wireless connectivity, audio quality, encryption and remote access. Investigators say the account also asked questions about concealing the hardware. Interestingly, the account submitted prompts mentioning a custom header and unique MAC address associated with the actual device discovered at OpenAI.The identity of the chatbot hasn't been publicly confirmed, either. There's no evidence in the public filing establishing that ChatGPT specifically was used, or that an AI model independently designed the device. (Image credit: Getty Images/Jakub Zerdzicki)OpenAI discovered the first device on September 30 during a security sweep. According to the Justice Department, it was attached to the back of a television in a conference room reserved for executives and executive assistants. The room regularly hosted confidential discussions about OpenAI's research, upcoming products, business strategy and relationships with government officials.The hardware was reportedly a Seeed XIAO ESP32-S3 Sense, a small development board equipped with a microphone, camera sensor and storage capability.Although the hardware itself is relatively inexpensive, OpenAI described the device as sophisticated because of its encryption, audio monitoring and data-transfer capabilities.Security personnel couldn't read the information stored on the device because its contents were encrypted. Investigators also reviewed surveillance footage that allegedly placed Alfozan near the conference room on multiple occasions.On October 5, security cameras reportedly captured him returning to the room and looking behind the television. That same day, OpenAI personnel discovered a second similar recording device inside a bag Alfozan had left unattended elsewhere in the office. The bag also contained identifying personal belongings.Alfozan was arrested on October 9 and appeared in federal court on October 13. He faces up to five years in prison and a $250,000 fine if convicted. The allegations remain unproven.Bottom lineIn the statement posted to X, OpenAI confirmed that its security team used the company's internal AI models to help understand the recording device.OpenAI has not said that AI independently identified the suspect. What makes this case unsettling is that ordinary AI tools may have helped someone research the technical and logistical details of a surveillance operation, making specialized information easier to obtain.This incident proves that a person no longer needs to work through technical documentation alone to investigate how hardware operates, troubleshoot configurations or understand unfamiliar equipment. AI assistants can help with those details in conversational language.Follow Tom's Guide on Google News and add us as a preferred source to get our up-to-date news, analysis, and reviews in your feeds. Subscribe to Tom's Guide on YouTube and follow us on TikTok. More from Tom's GuideTurn your PDFs into audiobooks with ChatGPT — here's the simple trickAI researcher warns cyberattacks could take down power, banking and water within a year — here's how to prepareCharacter.AI files for bankruptcy — what happens to your chats and AI companions now? Amanda Caswell is the AI Editor at Tom's Guide and one of today’s leading voices in AI and technology. A celebrated contributor to various news outlets, her sharp insights and relatable storytelling have earned her a loyal readership. Amanda’s work has been recognized with prestigious honors, including outstanding contribution to media.Known for her ability to bring clarity to even the most complex topics, Amanda seamlessly blends innovation and creativity, inspiring readers to embrace the power of AI and emerging technologies. As a certified prompt engineer, she continues to push the boundaries of how humans and AI can work together.Beyond her journalism career, Amanda is a long-distance runner and mom of three. She lives in New Jersey.