Western Universities Freely Share AI Research With Chinese Labs: Study

The report finds that such collaborations facilitated human rights abuses and mass surveillance of civilians in China.

Beijing’s information war against the United States is getting more sophisticated with the help of artificial intelligence. Getty Images

U.S. universities are “extensively” collaborating with Chinese labs on artificial intelligence research, which has also facilitated transfer of sensitive U.S. tech, mass surveillance of Chinese civilians, and human rights abuses in China, including what the U.S. government has designated a genocide of Uyghurs in the Xinjiang region, according to a study published by analysis firm Strategy Risks in partnership with the Human Rights Foundation.

The report, released on Dec. 8, focuses on case studies of two state-backed Chinese labs, which were found since 2020 to have coauthored almost 3,000 papers with overseas researchers, including those at elite U.S. universities receiving government grants.

It names leading universities, including the Massachusetts Institute of Technology, Stanford University, Harvard University, Princeton University, the University of Oxford, and University College London, while noting that there are more than 20 others that have funding from public institutions, as well as corporations such as Amazon. The organizations did not respond to inquiries from The Epoch Times by publication time.

“This report shows that Western universities and public funders are tightly linked to China’s state-priority AI laboratories through open but poorly scrutinized collaboration,” the report reads.

“Knowledge moves seamlessly across borders, even when the receiving institutions are inseparable from an authoritarian state.”

CCP-Backed Labs

The two labs in focus are the Zhejiang Lab and the Shanghai Artificial Intelligence Research Institute (SAIRI), two “lavishly resourced, Party-supervised” institutions embedded in the regime’s surveillance state system, according to the report.

Zhejiang Lab was founded in 2017 by the Zhejiang provincial government, the state-run Zhejiang University, and Alibaba. It received $1.25 billion in provincial government funding between 2021 and 2023, according to the report. It also partners with several state-backed institutions such as the China Electronics Technology Group Corp., which was sanctioned by the United States for building the Integrated Joint Operations Platform, the backbone of the Chinese regime’s social credit system.

SAIRI was established by Chinese state-run Shanghai Jiao Tong University in 2018 and since 2020 has been run by Lu Jun, a senior scientist from the sanctioned China Electronics Technology Group Corp. SAIRI researches computer vision, tracking, applied imaging, and other technologies that can be used to monitor groups. Its collaborators include Huawei and the Chinese Ministry of Public Security’s Third Research Institute, which is responsible for technical surveillance and digital forensics, according to the report.

In 2023, it helped develop the first AI-assisted shooting training system, and in 2024, it signed contracts with two companies—iFlytek and SenseTime—that build facial recognition and smart policing platforms. Both companies have been sanctioned for aiding the Chinese Communist Party’s genocide of Uyghurs in the Xinjiang region.

Western Universities and Taxpayer Funds

According to the report, the two labs published 11,000 papers between 2020 and August 2025, including about 3,000 with overseas coauthors.

“Zhejiang Lab and SAIRI are part of a wider system of state-run AI laboratories that are heavily funded, politically directed, and linked to defense and security organs,” the report reads. “Their scale and international reach show that problematic collaborations are not isolated cases but systemic features of China’s AI research architecture.”

A 2022 paper coauthored with Massachusetts Institute of Technology researchers with funding from the Pentagon’s Defense Advanced Research Projects Agency grant program examines advanced optical phase-shifting, which the report states is employed in imaging and sensing systems used in military reconnaissance, satellite surveillance, and biometric monitoring.

A 2024 paper coauthored with Carnegie Mellon University researchers, funded by the U.S. National Science Foundation and Office of Naval Research, examines multi-object tracking, which the analysts wrote is a “core enabler of automated surveillance.” The university did not respond to an inquiry from The Epoch Times by publication time.

A 2021 paper coauthored with Harvard University researchers studies advanced optics and computer vision.

“Zhejiang Lab’s collaborations with these universities show how Western technical expertise is integrated into research streams that align with China’s security and surveillance priorities,” the report states, noting the lab’s close ties with a foundational organization contributing to the genocide of Uyghurs.

“The risk arises from structural embedding with state defense actors accused of human rights violations, not from the scientific framing of the projects themselves.”

In 2021, SAIRI collaborated with several Western institutions on QDTrack, which analyzes and can track people or objects across videos, which the report notes is a core function of mass surveillance technology. In 2023, it partnered with several American universities to develop AlphaTracker, a tool for animal behavior analysis that also has identity recognition and object tracking functions.

These Western institutions often receive government grants, as noted publicly in the papers. The Chinese labs themselves are not listed as grant recipients but benefit through partnerships with the Western institutions, the report states, bypassing due diligence procedures.

“The picture that emerges is one in which Western public resources are repeatedly implicated in research collaborations with laboratories embedded in China’s security apparatus,” the report reads.

The report focuses on Chinese institutions with explicit ties to the Chinese regime or military, but researchers cautioned that even institutions not overtly backed by the Chinese regime have intelligence sharing obligations to the regime under several Chinese national security laws.

Ethics Gap

The analysts found that only a minority of AI ethics centers have condemned the Chinese regime’s human rights abuses: the UK-based Ada Lovelace Institute in June 2022, Stanford’s Institute for Human-Centered Artificial Intelligence in August 2023, and the AI Now Institute in 2019.

“Western AI ethics institutes have largely avoided confronting China’s use of AI for repression,” the report reads. “This silence does not reflect ignorance, but a gap in governance and responsibility.”

The report authors warned that continued silence may help “normalize” partnerships with Chinese state-backed institutions and complicity in contributing to the regime’s surveillance state. The report calls on ethics institutes—organizations founded to “interrogate the risks of advanced technologies”—to issue guidance on the human rights risks of partnering with Chinese labs.

The report authors called for reframing due diligence to include human rights, mandating transparency in research partnerships, putting safeguards in place for working with institutions linked to surveillance, and showing ethics leadership in the field. They also advised governments to revise policies to close the national security gap seen in these collaborations.

“Unless governments, universities, and ethics bodies adopt such measures, collaborations with Chinese state-priority laboratories will continue under current rules,” the report reads.