The most effective approach democracies can take to gain the upper hand in gleaning AI benefits is for them to cripple autocracies’ AI capabilities in certain areas. Successful AI requires three things: data, algorithms, and computing power. While autocracies’ lack of human rights means they enjoy an absolute advantage in data, they currently face the most challenges in the competition for computing power enabled by AI chips. While autocracies such as China have made some progress in developing their own AI chips, most AI applications run by major Chinese firms are empowered by foreign chips. This demonstrates how vulnerable autocracies such as China are to supply disruption in AI chips, meaning export controls imposed by democracies such as the United States will continue to be a powerful tool in disabling autocracies’ AI capabilities.
In the age of AI, the idea that technology could favor autocracies more than democracies seems an easy but desperate conclusion to draw. Indeed, it may even be the case. But if this outcome is realized, it would be democracies’ negligence—or outright inaction—that would be to blame. When asked if he agreed with the statement that “technology change today is strengthening authoritarianism relative to democracy,” Joseph Nye, former dean of Harvard Kennedy School, provided a powerful answer: “Only if we let that happen.”
Ciel Qi is a researcher focusing on China’s technological and economic development and foreign policy. She holds a master’s degree from Harvard University in religion, ethics, and politics and is an MA candidate in the Security Studies Program at Georgetown University, concentrating in technology and security. Her writing has previously appeared in publications including the TechCrunch Global Affairs Project.