Google’s China Ties in AI Research Eyed

Company collaborates with entities that conduct AI research with military applications

A security guard passing a booth of Google at the first China International Import Expo in Shanghai on Nov. 5, 2018. JOHANNES EISELE/AFP/Getty Images

Google’s connections to China have come under scrutiny after U.S. President Donald Trump in late July said his administration would look into allegations that Google was working with the Chinese government on projects that could threaten U.S. national security.

“There may or may not be National Security concerns with regard to Google and their relationship with China. If there is a problem, we will find out about it. I sincerely hope there is not!!!” Trump wrote in a July 26 tweet.

Google left the Chinese market in 2006, and its search engine is blocked by China’s “Great Firewall.” But the U.S. internet giant maintains a number of tech research projects in China, many of them focused on artificial intelligence (AI). The company has offices in Beijing, Guangzhou, Shanghai, and Shenzhen.

Google placed its emphasis on AI research in the country by appointing Li Fei-Fei, an AI researcher and professor at Stanford University, to start its AI China Center in Shanghai.

According to the company, the research center was established to encourage research collaborations with the country’s top AI and machine-learning experts.

While there is no information to suggest that the AI center conducted sensitive research that could have national security concerns, through scouring Chinese-language media reports and online sources, The Epoch Times found information about Google’s AI collaborations in China that indicates they have military applications.

In addition, Li, who left the company in September 2018, has extensive ties to Chinese AI research and academic circles. She is a member of a science and tech research forum that is supervised by Chinese authorities. In addition, mentors she has cited as important to her career are participants of the “Thousand Talents” program, an initiative started by Beijing to recruit top scientists and engineers from the West to work in China.

In recent months, the Thousand Talents program has come under U.S. scrutiny because of its potential in abetting academic espionage. Several Chinese individuals indicted on federal charges of stealing trade secrets were participants in the program.

Google and Li didn’t immediately respond to The Epoch Times’ requests for comment.

Research Project

On June 28, the official website of the Chinese Academy of Sciences, the country’s top state-run research institute, published an article entitled “Progress in the Basic Research on Human–Computer Interaction of Mobile Targets in Software.” The article stated that a top Chinese AI research scientist employed by Google, Zhai Shumin, came to Beijing to work with researchers at the academy on developing a “human-computer interaction” project, the results of which were published in a paper.

According to the paper, the researchers developed a kind of target selection-assisting technology, which increased the speed of users’ selection of moving targets by about 57 percent and accuracy by about 79 percent. At the same time, the team also used the model to predict the error rate of a moving target selection.

Upon publication, two unidentified Chinese researchers familiar with the project confirmed to Hong Kong-based newspaper South China Morning Post (SCMP) that the technology allows fighter pilots or air defense missile operators to select fast-moving targets quickly and accurately using a touch screen, and that the Chinese military’s most advanced stealth fighter jet, the J-20, has the opportunity to adopt the technology.

The Chinese Academy of Sciences stated in the initial version of its online press release announcing the research results: “The research … plays an important role in the interface design and technology development of human-computer interaction widely applicable in the military, medical, education, digital entertainment, and other fields.”

But after SCMP reported about the paper and sought comment from Google, the word “military” was deleted from the online press release.

But many Chinese media, such as Sina.com, which reprinted the press release, retained the word in their reporting.

Google confirmed to SCMP its involvement in the research paper, but denied a link to the military.

“This paper addresses a very general research question in user experience design of how people interact with moving items on a touch screen. ... This paper is simply not about military applications,” SCMP cited Google as saying. “Research like this is key to improving finger or stylus-based navigation in any app.”

Collaboration With Tsinghua

The Epoch Times previously reported that in June 2018, Google founded a new AI research body with Tsinghua University, one of China’s most prestigious schools.

Earlier that month, it was revealed that the university received significant funding from the Chinese military to work on a project aimed to advance the military’s AI capabilities.

China Education Daily, a state-owned newspaper run by the regime’s Ministry for Education, reported on June 8, 2018, that Tsinghua University received more than 100 million yuan ($14.53 million) from the Science and Technology Committee of China’s Central Military Commission—a Party organ that oversees the military—to work on an AI project for the military.

The project is tasked with researching and developing AI for human-machine combat teaming, the report said.

The report added that the work of the university’s military AI lab, called “Military Intelligent High-End Lab” and established in 2018, would be “guided by military needs” and would help build China into an advanced AI country.

The Chinese regime has set the development of AI as one of its top priorities, especially in “military-civilian integration,” or developing technology that can have both military and civilian applications. In July 2017, China’s State Council, a cabinet-like agency, published a detailed plan for China to become a “world leader” in AI by 2030. The plan aims to build a domestic AI industry worth $150 billion.

Li Fei-Fei

The Tsinghua lab collaboration was announced at a joint Google–Tsinghua symposium held in Beijing on June 28, 2018. Li Fei-Fei, then a vice president at Google Cloud, was in attendance. She was hired by Google in November 2016 to lead a new AI research unit.

Li was born in 1976 and immigrated to the United States with her parents at the age of 16, according to Chinese media reports that have noted her success as an AI expert.

She left Google in September 2018, when she announced she would resume teaching at Stanford, where she is director of the university’s Artificial Intelligence Laboratory and Vision and Learning Lab.

Li is involved with a number of science and tech organizations in China that have close ties to authorities in Beijing.

Li is a member of the Future Forum science committee. According to the forum’s official website, the organization was founded in Beijing in 2015 by leaders in the scientific, educational, internet, and investment fields to facilitate cross-disciplinary research. It is directed by the China Science and Technology Association (CAST) and supported by the Beijing Chaoyang District government.

CAST describes itself as a non-governmental organization, but also clearly states on its website that it “serves as a bridge that links the Communist Party of China and the Chinese government to the country’s science and technology community.”

Future Forum also lists “strategic cooperation media” partners such as People’s Daily, Xinhua, and The Paper, all Chinese state-run media.

Other members of the forum are descendants of or associated with former top Communist Party officials, commonly referred to as “princelings,” including Liu Lefei, son of Liu Yunshan, former member of the Politburo Standing Committee, the Party’s top decision-making body; Zhu Yunlai, son of former Premier Zhu Rongji; and Ma Xuezheng, a business executive at Boyu Capital, a private equity firm that was founded by Jiang Zhicheng, grandson of former Party leader Jiang Zemin.

In December 2017, at an “overseas talent exchange” conference held in Guangzhou City that was jointly organized by the European and American Alumni Association and the Chinese Ministry of Education, Li was selected as one of 50 top Chinese who studied abroad.

The European and American Alumni Association was founded by late imperial intellectuals in Beijing in October 1913. It was originally a group for Chinese intellectuals who studied abroad.

After 1949, the association was taken over by the communist regime and became an important tool for the Chinese Communist Party’s United Front Work Department, an agency tasked with spreading the regime’s agenda at home and abroad.

An August 2016 report by state-run media Xinhua noted that Beijing authorities issued a notice on how to better build Party allegiance in the Alumni Association, explaining that it’s an organization “directed by the Party” as a “united front mass organization.”

Many former Party leaders, top researchers at state-run institutes, and university presidents are members of the association.

Li’s Mentors

In July 2017, Li delivered a speech before the Conference on Computer Vision and Pattern Recognition, where she thanked her mentors for their support.

Li said she got her first teaching job at the University of Illinois at Urbana-Champaign with the help of Thomas S. Huang, or Huang Xutao. A naturalized U.S. citizen and research professor in electrical and computer engineering at the university, Huang in 2002 was appointed a foreign academician of the state-run Chinese Academy of Sciences. In 2012, Huang was recruited under the Thousand Talents program by a research institute in Chongqing City in southwestern China.

Li also mentioned Li Kai, a professor at Princeton University, during her speech. She explained how she received his help while ​​organizing the international AI imaging competition ImageNet.

She said that Li Kai allowed her to use machines in his lab, and made the competition possible.

In November 2017, Li Kai was also elected a foreign academician, of the Chinese Academy of Engineering. He is a lecture professor at Tsinghua University and also a Thousand Talents Program recruit.

They haven’t been accused of any wrongdoing. But U.S. authorities are on the alert about China’s recruitment programs. At a congressional hearing in December 2018, Bill Priestap, then-assistant director of the counterintelligence division at the FBI, said that China’s recruitment programs, such as the Thousand Talents, were “brain gain” programs that “encourage theft of intellectual property from U.S. institutions.”

Earlier this year, the United States’ National Institutes of Health published a report about the risks of state-sponsored programs such as Thousand Talents for presenting conflicts of interest, as scientists may be recruited by foreign countries to conduct research that could be funded with U.S. federal grants.

Li Fei-Fei didn’t immediately respond to questions about her relationship to Li Kai and Huang, nor her affiliation with the alumni association.