In July 2018, Xiaolang Zhang made his way to the San Jose, California airport with intentions to board a flight to Beijing. However, his journey was abruptly halted at Terminal B when federal agents, acting on a tip from Apple's security team, arrested and charged the former Apple employee with stealing trade secrets associated with the company's autonomous-driving program.
It marked another engagement in the ongoing clandestine conflict between the United States and China for supremacy in artificial intelligence—a strategic competition where both nations strive for any advantage to secure a leading position in harnessing a technology capable of reshaping economies, geopolitics, and warfare.
The Federal Bureau of Investigation (FBI) has identified artificial intelligence as a critical U.S. technology requiring safeguarding. Similarly, China has designated it as a technology for breakthroughs by its scientists by 2025. While China is recognized for its formidable AI capabilities, recent warnings from U.S. intelligence authorities extend beyond concerns about intellectual property theft.
Rather than solely focusing on pilfering trade secrets, the FBI and other agencies express concerns that China might exploit AI to collect and amass data on Americans at an unprecedented scale. FBI Director Christopher Wray, in a press conference in Silicon Valley earlier this year, suggested that China, known for various significant data thefts, could use artificial intelligence as an "amplifier" to support further hacking operations.
"Now, they are leveraging AI to enhance their already extensive hacking operations, utilizing our own technology against us," stated Wray.
China has consistently denied involvement in hacking into U.S. networks. Wang Wenbin, a spokesman for the Chinese Foreign Ministry, asserted this summer that the U.S. was the "biggest hacking empire and global cyber thief" in the world, countering allegations of Beijing hacking into the unclassified email systems of high-ranking Biden administration officials. The Chinese Embassy in Washington did not respond to requests for comment.
In recent times, the FBI's focus on safeguarding American innovations has shifted more towards manufacturers of powerful chips capable of processing artificial intelligence programs, rather than the artificial intelligence companies themselves. Former U.S. officials noted that even if insiders or hackers managed to steal the algorithms supporting an advanced system today, the rapid pace of technological advancements could render that system obsolete in six months, overtaken by larger strides made by other engineers.
In 2022, Applied Materials, a chip-manufacturing technology supplier, filed a lawsuit against Mattson Technology, a China-owned competitor. The lawsuit alleged that a former Applied engineer stole trade secrets before joining Mattson, attracting the interest of federal prosecutors, though no criminal charges have been filed.
Mattson, based in Fremont, Calif., acquired by an investment arm of the city of Beijing in 2016, asserts that no federal agency has contacted them, and there is no evidence of the stolen Applied information being used. In a twist, Mattson sued Applied in November, claiming that engineers at Applied had applied for patents using intellectual property developed while working at Mattson.
Growing concerns about how China could leverage AI prompted FBI directors and Western intelligence leaders to meet with technology experts in October. Executives in the AI technology sector are also worried about their secrets reaching China. OpenAI reached out to the FBI after suspicions arose regarding a former employee taking company secrets to China; however, the employee was later exonerated.
For years, U.S. intelligence analysts have expressed worry about China's long-term gains from amassing vast troves of hacked personal information. China's involvement in significant data breaches has been linked to numerous incidents, including those affecting Marriott International, Equifax, and Anthem. China's adeptness at stealing private information, in the billions, has raised concerns that the amassed data might be too vast for human analysis but could be effectively processed using AI.
Microsoft believes China is already using its AI capabilities to analyze extensive datasets, citing evidence from targeted attacks. The fear is that AI will be employed to refine and enhance targeting in cyber operations. However, the issue of China stealing AI technology is not directly addressed.
In a 2018 case, a former Apple employee, Zhang, pleaded guilty to stealing trade secrets related to Apple's autonomous-driving program. Set to be sentenced in February, Zhang's plea agreement is sealed, and Apple declined to comment.
U.S. authorities believe Chinese intelligence operatives are using AI to correlate stolen sensitive information, including fingerprints, foreign contacts, financial details, and medical records, to locate and track undercover U.S. spies and officials with security clearances. The potential national security threat is underscored by China's ability to harness AI to compile extensive dossiers on Americans.
Despite concerns about the weaponization of AI, executives, including Microsoft's Brad Smith, emphasize that AI can be used as a potent defensive shield against cyber threats if managed effectively. The goal is to use AI to detect and mitigate attacks rather than as an offensive weapon.