In what is perhaps the tensest scene in Stanley Kubrick’s 2001: A Space Odyssey, the spaceship’s onboard artificial intelligence system HAL-9000 reads the lips of hiding crew members. Discovering their plans to turn it off, the system turns against all humans onboard for the sake of protecting the mission.
What was once science fiction is now inching closer to reality thanks to artificial intelligence—but in the wrong hands it could open up terrifying possibilities. In 2018, Chinese researchers crafted a new dataset and surveillance application benchmarks for lip-reading in real-world settings. Though this tech was meant to aid the hearing-impaired, the Chinese are likely to use it to broaden their already expansive surveillance state. And any serious surveillance tech in the hands of the Chinese is bad news for everyone else.
Using a commercial technology for surveillance or military purposes is known as “dual use,” and the People’s Liberation Army (PLA), China’s armed forces, leads in this field. Chinese universities are closely connected to military activities and aligned with the Communist Party’s ambitions—and they’re also among the world’s leaders in AI research. The global AI community has already voiced concern over this dual use and created an action plan to enhance global security. Nonetheless, the United States must take the helm in setting the technology’s global ethical standards. Otherwise tyrannical leaders will be empowered to use AI for malicious purposes.
Before that can happen, Congress has to understand what it’s up against. And it doesn’t.
Congressional expertise has declined significantly since the 1990s. The Office of Technology Assessment (OTA) was defunded in 1995 and the Government Accountability Office, meant to fill that role since 2007, has since experienced a decline in staffing. This glaring lack of knowhow is worrying. The number of congressional members with no technical background far exceeds those with STEM qualifications, a problem that’s persisted despite a midterm increase in such STEM representation. Without sufficient technical knowledge, Congress will be unable to fully anticipate the ways the Chinese can wield their vast tech power—and will be left reeling when they do.
Misconceptions about how strong the Chinese tech sector really is have led the U.S. to view them as mere copycats rather than innovators. The Trump administration’s response to China has focused on the pervasive problems of supply chain dominance and the intellectual property theft of American innovation. This has allowed the PLA to quietly increase its legitimate research capacity. At least 2,500 PLA military scientists and engineers have developed relations with free-world institutions since 2007, according to research by the Australian Strategic Policy Institute. And these military affiliations are often obscured: they’ve allowed PLA scientists to co-author over 2,500 studies abroad since the number of initiatives rose in 2013. As these publications grow, so will Chinese influence. The same scientists with the newfound ability to direct international AI research are also committed to the Communist Party.
Luckily, the United States remains the dominant voice in AI. It’s home to the most respected research institutions and innovative companies in the world. But this edge is close to gone, and its maintenance requires political leadership. Current proposals, such as export controls to limit the sale of advanced algorithms to nations like China, only protect—rather than direct—research, and fail to really confront China’s growing AI power. As Beijing increases its research capacity and nations become dependent on Chinese technology, America will need a more proactive approach.
The U.S. should exert greater influence on the direction of global research, ensuring that America remains dominant in basic research and that global AI progress respects individual freedoms. This will require greater commitments to science and technology funding and a regulatory environment that allows for safe yet rapid innovation. America’s policymakers must increase their technical expertise, articulate ethical standards, and reduce the dependence of researchers on Chinese military collaborations.
Initiatives like Future Congress, which push to refund the OTA and increase congressional technical expertise, are urging the United States to keep its rightful place as leader in science and technology. The U.S. should also champion the excellent G7 proposal to create an International Panel on Artificial Intelligence to assess AI progress and the risks posed by advancement.
If Washington fails to support cutting-edge AI innovation and the standards that support its proper use, that responsibility will fall into the eager hands of the communist Chinese. And if that happens, free nations will be in for a world of trouble.
Ryan Khurana is executive director of the Institute for Advancing Prosperity and a Technology Policy Fellow at Young Voices.