US Faces Escalating Chinese IP Theft Targeting AI and Advanced Semiconductor Technologies

Intellectual property theft by Chinese entities has long challenged the United States and its allies, but recent developments indicate a troubling intensification targeting artificial intelligence (AI) and advanced semiconductor chips. A new White House report reveals that Chinese Communist Party (CCP)-backed groups are conducting large-scale theft operations against American AI research labs, extracting proprietary information through sophisticated methods.

Michael Kratsios, director of the White House Office of Science and Technology Policy, detailed in a memo that these coordinated campaigns employ tens of thousands of proxy accounts to evade detection and use jailbreaking techniques to access confidential AI model data. This systematic exploitation leverages American innovation to replicate AI capabilities.

The U.S. State Department has issued a global warning concerning Chinese AI theft, particularly implicating the Chinese startup DeepSeek. Kratsios explained that Chinese firms primarily use "distillation attacks," a process by which they utilize American AI models such as Claude or Gemini to generate datasets that mimic the behavior of these systems. This enables the creation of similar AI models at significantly reduced costs.

In parallel, China is aggressively attempting to smuggle advanced semiconductor chips essential for AI processing out of the United States. The Trump administration prioritized export controls to prevent the CCP from acquiring cutting-edge chips, aiming to maintain a technological edge.

During a recent Congressional Select Committee on China hearing, Chairman John Moolenaar estimated that China has successfully smuggled tens of thousands, possibly hundreds of thousands, of AI chips from the U.S. He cited a landmark case involving $2.5 billion worth of products, marking the largest export control violation in U.S. history. The case involved three employees of Super Micro Computer—Yih-Shyan Liaw, Ruei-Tsang Chang, and Ting-Wei Sun—who were charged with illegally diverting Nvidia AI chips to China.

According to the Department of Justice, the conspirators knowingly violated export laws, motivated by promises of multimillion-dollar payouts. They concealed their activities by listing fictitious Thai buyers acting as intermediaries for Chinese companies. All three were charged in March 2024.

Another smuggling scheme involved routing chips through a Taiwanese company, deceiving U.S. inspectors by substituting genuine chips with physical replicas. The smugglers employed elaborate tactics, including using a hair dryer to alter label packaging and placing counterfeit chips in original boxes to evade detection.

In a separate 2024 case, Google employee Linwei Ding was convicted on multiple counts of economic espionage and theft of trade secrets after stealing over 2,000 pages of sensitive information. Ding sold this data to Chinese firms intending to help them match U.S. computing capabilities. While still employed by Google, Ding applied for a Chief Technology Officer position at a Chinese company and founded a startup in Shanghai claiming to develop state-of-the-art supercomputers.

A former Chinese military intelligence officer reviewed Ding’s case, noting that his approach reflected classic CCP intelligence tactics of operating openly to avoid suspicion. "The darkest place is under the candlestick," the officer remarked, emphasizing Ding’s use of Google’s own tools to blend in while transferring sensitive data.

Beyond advanced chips, China is also striving to dominate the market for less sophisticated semiconductors. AI expert Dmitri Alperovich described this as a "new Strait of Hormuz," warning of a strategic threat if Beijing controls the supply of affordable chips critical to global demand. He noted that China employs state-backed, non-market policies similar to those used in steel, solar, and electric vehicle sectors, aiming to capture over half of the global AI chip market by 2030.

Understanding China’s relentless pursuit of American innovation requires recognizing that technological competition with the U.S. is central to the CCP’s political identity. Dr. Shufen Yóulǎn, a retired economist and former lecturer at CCP schools, explained that failure to achieve breakthroughs would undermine the Party’s leadership. Xi Jinping’s authority has already been challenged by military commanders who criticized his aggressive stance toward the West.

Dr. Yóulǎn stated, "The party boasted about matching progress for years, but now even that illusion is fading. Haunted by the fear of losing its propaganda edge, the party manipulates or steals American intellectual property and claims it as a victory for Xi’s New Era."

German diplomat and China expert Professor Jürgen Everhart summarized the CCP’s modus operandi succinctly: "Deception is the CCP’s lifeblood."

In response, four bills currently advancing through Congress aim to counter these threats. The MATCH Act seeks to coordinate allied bans on AI technology exports. The AI OVERWATCH Act would require export licenses for high-risk countries. The SCALE Act intends to maintain U.S. technological superiority by restricting exports. Lastly, the Remote Access Security Act proposes controlling advanced AI technology by offering it exclusively via cloud platforms, potentially disrupting CCP access.

A Chinese dissident familiar with these legislative efforts remarked that such measures "would shatter the CCP’s backbone."

As AI technology rapidly evolves, the United States and global community face a pivotal moment. The outcome of this technological rivalry will shape which great power controls the future being constructed by artificial intelligence.