OpenAI Secrets to China? Suspected Data Leak Sparks FBI Involvement

The use of AI on accumulated hacking data could lead to targeting…
Concerns grow over Chinese intelligence agent ‘Longtail’ spy activities

OpenAI, the developer of ChatGPT, recently reported to the FBI after forensic investigations of all employee laptops raised suspicions that a certain employee had leaked company secrets to China. While the employee in question has been cleared of all charges, concerns about Chinese leaks of confidential information are escalating within both AI companies and U.S. intelligence authorities.

On the 25th, the Wall Street Journal (WSJ) highlighted that China’s persistent and covert spy operations, long feared by U.S. intelligence analysts, are now becoming a stark reality. In the field of cybersecurity, experts utilize Long-tail analysis to pinpoint subtle signals from highly skilled attackers who successfully elude detection. Growing evidence suggests that China is amassing extensive personal data on U.S. officials and corporate executives through cyberattacks and utilizing AI capabilities to refine its targeting strategies.

U.S. authorities claim that Chinese intelligence agents are tracking and identifying U.S. spies and managers with security clearance by linking sensitive information such as fingerprints, overseas contacts, financial debts, and personal medical records from databases they have stolen over the years. For instance, passport information stolen in the Marriott hack could assist spies in monitoring government officials’ travels.

Brad Smith, Microsoft’s president, shared his observations on the evolving landscape of cybersecurity with the Wall Street Journal. He noted that the critical question in the past revolved around the ability of various entities, including China, to utilize machine learning and AI to gather data from hacks and employ it in targeting strategies. Smith highlighted that there has been tangible evidence of this practice over the last two years. He further emphasized the necessity to assume that China will persistently utilize AI to enhance and refine its targeting techniques.

The FBI and other agencies believe that China is not just stealing corporate trade secrets but is using AI to collect and stockpile data on Americans on an unprecedented scale. At a press conference earlier this year in Silicon Valley, FBI Director Christopher Wray issued a warning, highlighting that China has been implicated in numerous cases of personal data theft over the years. He also emphasized that AI could be an “amplifier” to facilitate hacking operations.

Glenn Gerstell, a former legal advisor to the U.S. National Security Agency, said, “China can harness AI to build a dossier on virtually every American, with details ranging from their health records to credit cards and from passport numbers to the names and addresses of their parents and children.” He explained that adding hundreds of thousands of hackers working for the Chinese government to this equation could pose a potential threat to U.S. national security.

During the past decade, China has been implicated in hacking incidents that encompass billions of customer records from notable organizations, including Marriott International, credit agency Equifax, health insurer Anthem (now Elevance Health), and more than 20 million personnel files of current and former U.S. officials and their families from the Office of Personnel Management. One prominent instance was the 2021 attack, attributed to China, which targeted tens of thousands of servers running Microsoft’s email software.

By. Hee Jeong Kim