An Associated Press investigation uncovers the stark reality that the US government has, for decades, allowed and even aided American tech companies in providing surveillance technology to China. This deeply contradicts public rhetoric on national security and human rights, highlighting how legislative loopholes and intense lobbying from tech giants have consistently undermined efforts to restrict such sales, empowering Beijing’s control over its populace.
In a striking revelation, an Associated Press investigation has brought to light a profound contradiction in U.S. foreign policy: for decades, while publicly cautioning against national security threats and human rights abuses, the U.S. government has consistently allowed, and at times actively assisted, American firms in selling critical technology to Chinese police, government agencies, and surveillance companies. This extensive complicity spans five Republican and Democratic administrations, fostering the growth of China’s pervasive surveillance state.
The investigation reveals a pattern of legislative failures and powerful lobbying that has circumvented efforts to curb these sales. Lawmakers, for instance, have attempted four times since September 2024 to close a significant loophole allowing China to bypass export bans on powerful American AI chips. Instead of direct purchases, Chinese entities rent these chips through U.S. cloud services, a workaround that has consistently survived congressional scrutiny due to intense lobbying from over 100 tech company representatives and their trade associations.
The Unseen Influence of Tech Lobbying
The tech industry’s immense wealth and political power are central to understanding this ongoing issue. An AP analysis of lobbying disclosures shows that U.S. tech and telecom companies, along with their associated trade groups, have spent hundreds of millions of dollars over the last two decades influencing legislation related to China trade. This financial might has often overshadowed national security and human rights concerns, as evidenced by the repeated failure of bills designed to tighten export controls.
Tech companies argue that further restrictions would be counterproductive, pushing China to develop its own domestic supply chain and strengthening its global standing in the race for artificial intelligence dominance. For example, Nvidia stated that “continuing to ban U.S. computing from commercial markets only benefits foreign competition and undercuts President Trump’s efforts to create jobs, reduce the trade deficit, and grow the economy.” Nvidia also clarified that it does not develop surveillance systems or software, nor does it work with Chinese police or design its H20 AI chip for police surveillance.
Even more concerning are the direct government dealings. In August 2024, former President Trump announced a deal with chipmakers Nvidia and AMD to lift export controls on advanced chip sales to China. This agreement granted the U.S. government a 15% share of the revenue, despite stark warnings from national security experts that these chips could easily end up in the hands of Chinese military and intelligence services. That same month, the U.S. government acquired a 10 percent stake in Intel, valued at approximately $11 billion, further intertwining U.S. economic interests with tech exports to China. Intel affirmed its adherence to export control policies and described the government’s investment as “passive ownership, with no board representation, governance or information rights.”
Persistent Loopholes Empowering China’s Surveillance Apparatus
The investigation highlights several key loopholes that China has exploited:
Cloud Services Loophole: Chinese companies, officially barred from directly acquiring cutting-edge chips, circumvent these restrictions by renting computing power through U.S. cloud platforms like Microsoft Azure and Amazon Web Services (AWS) overseas. Both Microsoft and AWS even advertise their capacity to store video surveillance footage for Chinese customers. This includes state-owned tech companies like SDIC Contech and government-backed research institutes such as the Shanghai Qi Zhi Institute, which sought access to advanced analytics and OpenAI cloud services.
Sanctions Bypass: Even sanctioned Chinese surveillance giants, Dahua and Hikvision (sanctioned over human rights abuses in Xinjiang in 2019), leverage AWS to offer networked surveillance solutions abroad. Hikvision explicitly markets its “HikCentral” video surveillance platform for deployment on Azure, demonstrating how deeply U.S. cloud infrastructure is enmeshed with Chinese surveillance capabilities, despite sanctions. Microsoft, however, denied direct service provision to Hikvision or partnerships for such services, and OpenAI stated its adherence to Microsoft’s policies, not supporting China access to its models.
Outdated Tiananmen Restrictions: Restrictions enacted after the 1989 Tiananmen Massacre primarily targeted low-tech “crime control and detection” equipment (restraints, helmets, batons), neglecting advanced technologies like security cameras, surveillance drives, and facial recognition systems. Numerous bipartisan legislative attempts in 2006, 2007, 2009, 2011, and 2013 to update these restrictions and close the loophole all failed.
The U.S. Commercial Service, the Commerce Department’s export-promotion arm, actively facilitated these connections for over a decade. Archived webpages detail how it connected U.S. vendors with Chinese security agencies, promoted U.S. participation in policing trade shows (even those initiated by the Chinese Communist Party), and hosted webinars on selling surveillance tools to China’s public sector. This proactive support, across multiple administrations, underscores a systemic failure to align trade policy with stated human rights and national security objectives.
Voices of Disappointment and the Human Cost
For activists like Zhou Fengsuo, a student leader during the 1989 Tiananmen protests and now a U.S. citizen, this situation is a “strategic failure by the United States.” Zhou, who testified before Congress in 2024 calling for an investigation into American tech involvement in Chinese surveillance, believes U.S. companies are “driven by profit” and are ignoring their role in enabling Beijing to surveil and censor its own population. His testimony, available through official congressional records, provides a stark reminder of the historical context of these issues: Zhou Fengsuo’s written testimony to Congress highlights the long-standing concerns.
The human toll is particularly evident in the experience of Gulbahar Haitiwaji, an ethnic Uyghur detained in Xinjiang internment camps for over two years. Her arrest was facilitated by policing systems based on U.S. technology. Haitiwaji, who endured constant surveillance even in the toilet, testified to Congress in 2023, imploring the U.S. government to “stop American companies from continuing to be complicit in surveilling our people.” Her powerful plea underscores the direct impact of these tech sales on persecuted communities. Haitiwaji’s detailed testimony is publicly available, detailing her harrowing experience.
Politicians on both sides of the aisle echo this sentiment, blaming the failures on the immense financial and political leverage of tech companies. New Jersey GOP Representative Chris Smith stated, “I think we’ve been naive or complicit in the extreme,” warning that the U.S. is “selling and conveying to a malevolent power the ability to destroy us and destroy like-minded Western democracies.” Senator Ron Wyden (D-Oregon) succinctly added, “What do all those companies all have in common? A big wallet. That is as much as anything is what’s behind the fact we haven’t made as much progress.”
A History of Missed Opportunities and Future Implications
Attempts to regulate these exports have a long and unsuccessful history. In 2008, the Department of Commerce sought public comment on including “biometric devices” and “integrated security systems” under controlled exports, but the initiative stalled. Later, in 2014 and 2015, the government tried to strengthen controls on surveillance products, with most efforts falling through. As recently as 2024, there was an attempt to restrict face-recognition systems and broaden the list of military, police, and intelligence end-users barred from receiving U.S. goods, which also failed. These repeated failures highlight the systemic challenges in regulating rapidly evolving technologies against powerful industry interests. The Federal Register notice from 2008 offers insight into the Commerce Department’s initial efforts.
Despite congressional efforts like the 2018 Export Control Reform Act, which aimed to grant Commerce more authority over advanced technologies, and the Trump administration’s sanctions on some Chinese surveillance firms in 2019 and 2020, sales continued. While the pace slowed, and references to working with Chinese police vanished from Commerce Department reports, the underlying flow of technology persisted.
In 2021, President Joe Biden issued an executive order designating Chinese surveillance tech companies as “unusual and extraordinary threats” due to their role in human rights abuses. His administration subsequently drafted extensive rules for exporting advanced AI chips. The Commerce Department also revived a draft rule to restrict facial recognition and other mass surveillance tools from reaching problematic entities in China. However, this initiative also met with significant pushback from Washington lobbyists and industry groups, including the Chamber of Commerce, who argued it would “slow business considerably” and lead to customer loss, ultimately causing the proposed rule to stall out like so many before it.
The continued enablement of China’s surveillance capabilities by U.S. tech and government policies presents a complex challenge. It underscores a tension between economic interests, technological leadership, and fundamental human rights. As China continues to bolster its homegrown industry, evidenced by its $20.7 billion purchase of chipmaking equipment from U.S. companies in 2024, the long-term impact on global tech dominance and ethical responsibility remains a critical discussion for the tech community.
