Microsoft's top scientist is warning Congress about the dangers of Chinese artificial intelligence even as his company deepens its collaboration with China.
Dr. Eric Horvitz this year reported to Congress that "China's domestic use of AI is a chilling precedent for anyone around the world who cherishes individual liberty." The report was the result of Horvitz's work on a panel tasked with advising Congress on artificial intelligence. The report comes as Microsoft CEO Brad Smith is meeting with top Chinese officials to discuss deepening ties in digital technology and Microsoft's work in China.
Microsoft is one of many companies that capitulate to China in exchange for access to the country's lucrative market. The National Basketball Association supports political activism among its players and coaches but remains silent on China's human rights abuses and suppression of free speech. Apple allows Chinese officials to physically monitor its facilities and data centers in the country. Tech companies have also faced pressure from within their ranks to downplay Chinese human rights abuses. Apple employees have called on the company to condemn Israel for its actions in the Gaza Strip but not China for its use of slave labor.
Microsoft allows China to heavily censor LinkedIn, a Microsoft subsidiary. LinkedIn creates both a Western and a Chinese profile for Western users and shutters the Chinese versions when users post phrases banned by the Chinese government. In March, LinkedIn took a step further and banned a China critic based in the United Kingdom.
Smith continued his engagement with the Chinese government in a meeting last week with Xiao Yaqing, a top regulator at China's Ministry of Industry and Information Technology. According to Xiao, the two discussed expanding Microsoft's work with China.
Horvitz sounded the alarm about China as part of his work for the National Security Commission on Artificial Intelligence. The commission's report repeatedly warns that China's AI use is a threat to both liberty and American national security. China's "employment of AI as a tool of repression and surveillance—at home and, increasingly, abroad—is a powerful counterpoint to how we believe AI should be used," the report's authors say.
The report also highlights the dangers of China's use of artificial intelligence to more effectively leverage facial recognition technology. The authors write that China's use of AI to repress its Uyghur minority "foreshadows how authoritarian regimes will use AI systems to facilitate censorship, track the physical movements and digital activities of their citizens, and stifle dissent." Microsoft has expanded data centers in China for its Azure cloud service, which includes a facial recognition software package.
China in March hacked hundreds of thousands of Microsoft servers, an attack that some scholars believe was an attempt to collect data to train AI models. Microsoft responded to the attacks by expanding its business in China. The Biden administration refused to sanction Chinese actors for the attack.
Microsoft did not respond to requests for comment.