The question of ‘patriotism’ in U.S.-China tech collaboration

BEIJING —  In July, billionaire investor and entrepreneur Peter Thiel called for an FBI and CIA investigation of Google, saying the company was “treasonous” for allegedly working with the Chinese military instead of the U.S. military.

Thiel’s accusations were rejected by Treasury Secretary Steven T. Mnuchin, who said he and President Trump had met with Google CEO Sundar Pichai at the White House and found no evidence of Google working with the Chinese government or military.

Thiel has been criticized for trying to stoke U.S.-China trade war tensions for his own business advantage. But his comments were also indicative of a changing landscape in U.S.-China tech collaboration, as growing distrust between the two superpowers causes American companies and institutions to reconsider what safeguards should be in place to ensure that working with Chinese partners does not impinge on national security or human rights.

Several U.S. lawmakers — both Democrats and Republicans — recently called for an examination of U.S. funding that enables Chinese government surveillance of Muslim minorities in Xinjiang province. Scientists have been criticized for knowingly or unknowingly using Chinese data sets that draw on biometric data forcefully collected from minorities.

On the national security front, researchers who help China with applications for artificial intelligence have been accused of enabling Chinese military advancement. Thiel, whose company Palantir works extensively with the Department of Defense, says the “patriotic” thing to do is for tech to serve U.S. military interests.

But many experts and academics disagree. They say the risks of Chinese tech collaboration are serious, but that ethical questions are better centered on standards of transparency and consent than on notions of loyalty and patriotism.

The most critical difference between the U.S. and Chinese systems, they say, is that American companies and individuals have the freedom to choose whether they want to work with the government.

Framing tech in terms of “patriotism” is problematic, they say, because it suggests that companies or individuals exercising that freedom to reject U.S. military partnership are somehow unethical, when in fact they are demonstrating a strength of the American system.

“It is important for individuals to have recourse against what government asks us to do. That’s the most American thing I can think of,” said Kara Frederick, an analyst at the Washington, D.C.-based think tank Center for a New American Security who previously worked on a global security team at Facebook.

Democracy on its own is not a sufficient safeguard against privacy violations and misuse of data, as shown by the scandal involving data-mining firm Cambridge Analytica, which secretly harvested information acquired from Facebook and used it to influence the 2016 U.S. presidential race. But at least the rule of law and protection of freedoms of speech can provide opportunity for public debate and accountability, according to analysts.

Whether ethical standards can hold in U.S.-China tech collaboration is even less certain because of China’s authoritarian political structure.

Palantir co-founder Joe Lonsdale has referred to these differences as a clash of “two very different cultures.” When Google supports AI research with Chinese universities, it essentially serves the Chinese government, he said in a CNBC interview, because “everything in China is the government.”

“If you work in China, you do work with the government,” Lonsdale said. “That’s how China works.”

Frederick agreed that tech companies should reexamine their Chinese research partnerships. According to a National Intelligence Law passed in 2017, the Chinese government can command any individual or organization to hand over information and technology in the name of national security.

“You can’t just make something and naively believe that when you put it out into the world, people are always good and going to use it for good things,” Frederick said.

But it’s also oversimplification to conflate Chinese people and institutions with the Chinese government, or to label all collaboration as treasonous support for the Chinese military, said Elsa Kania, a technology and security analyst at the Center for a New American Security.

“The fact that Google is working on AI in China doesn’t mean they’re supporting the Chinese military directly,” Kania said, adding that even many Chinese companies are not “excited to work with the military.”

“If you’re a company fiercely competing for commercial advantage, working with the [People’s Liberation Army] is not your best interest,” she said.

Chinese tech employees told The Times they rarely question ethical implications of their work, but also that politics played a minimal role in their career decisions.

“It’s not like in the U.S., where people try to stay away from the government. Here, association with the government is somehow more honorable,” said Cathy, 26, an employee of a major tech company in Beijing who spoke on the condition of withholding her full name to avoid any repercussions from her employer.

Friends working for Huawei had thought of leaving because of the unhealthy overwork culture, she said, but changed their minds once Huawei executive Meng Wanzhou was detained in Canada at the U.S. government’s request.

“In a moment like that, they feel like what they’re doing is to some extent more worthwhile. You’re fighting for a bigger cause because you’re not only standing for the company but also for your country as a whole,” she said.

At the same time, Cathy said her career decisions were driven mostly by her goals for personal advancement. “I wouldn’t work for a company just because it gets bullied abroad,” she laughed.

James, 27, an employee in a microchip assembly factory in Shenzhen who also asked that his full name be withheld, said many young Chinese are struggling just to find employment.

“What I care about day to day is whether the traffic is good, whether I have work, whether the city will be a bit more convenient,” he said. “You talk about politics for half an hour, what do you get from it? Everything is status quo.”

Most young Chinese tech workers don’t have the luxury of thinking about the ethics of who they work for, James said. It’s enough of an economic challenge for people to move to the big cities where major tech companies are located.

“If I could work in one of those big tech companies, I would be absolutely thrilled,” he said. “I would go to whichever company I can get into.”

In the U.S., thousands of employees of major companies have protested, some even resigning on ethical grounds — most pertinently at Google, where employees protested Project Dragonfly, the company’s plan to create a censored search engine for China, and Project Maven, a partnership with the Department of Defense in AI that could be used for drone strikes.

Jack Poulson, an ex-Google researcher who resigned over Project Dragonfly and now runs a nonprofit promoting tech industry accountability, said that tech workers and academics should demand full transparency on the human rights implications of any project they’re tasked with.

“I don’t think it’s fair that just because you’re a tech company, you contribute to loss of life,” Poulson said. “As a civilian, you have the right not to contribute to any weapons system or any deprivation of rights or safety.”

Thiel’s statement was framed as a warning about China but implied that tech workers should instead serve the U.S. military, Poulson said.
But tech ethics is not about tying tech to one or another country’s unchecked military interests, Frederick said.

“It’s not just a hard guns and drones and offensive situation. It’s a battle for how this tech is used, to help or hurt people,” she said. “We’re not going to use virtual scanning for specific minorities and religious groups. We need to say that.”

Huang Yasheng, an MIT professor who heads the business school’s China Lab, said that civil-military fusion should be “taken as a given” in China and treated with appropriate safeguards, but should not stop engagement.

Disengagement would be particularly counterproductive in AI, Huang said, where China is technologically ahead of the United States.

“The way to advance AI is to make data bigger, which requires collaboration and sharing,” Huang said, adding that involving Chinese scientists in ethics discussions about ethics is a positive strategy.

“Maybe in the end, we decide to split. It’s like a divorce,” Huang said. “Do you go to divorce after a few instances of conflict, or to some intermediate step first?”

“If the two countries split, that’s not a tech future that is good for humanity, because that would enable China to do even worse than they’re doing now. We should collaborate as a way to passively influence their behavior.”