In the complex interplay between technological innovation and authoritarian rule, an MIT study sheds light on the dynamic development of AI-based facial recognition video surveillance technologies in China. This intriguing relationship underscores how innovation can serve the needs of authoritarian regimes and vice versa.
Historically, authoritarian governments have often resisted technological innovations, perceiving them as disruptive forces that threaten their hold on power. However, avoiding innovation can lead to stagnation and economic decline. Examining the specific case of China, researchers from the Massachusetts Institute of Technology (MIT), the National Science Foundation, Harvard University, and British institutions found that China leverages innovation to maintain control over dissident populations while driving economic development.
In China, facial recognition technologies powered by artificial intelligence have been widely deployed by the government for years, primarily as a means to repress and quell protests. The study revealed that in areas with heightened political tension, there is a significant increase in government procurement of AI-enhanced population identification systems.
While the regime heavily invests in AI for political control, it inadvertently stimulates innovation that can be exported globally. This phenomenon is aptly termed “AI-tocracy” — where advancements in AI serve both to suppress dissent and bolster the nation’s innovative capabilities.
To substantiate their research, the team utilized data from the GDELT project (Global Database of Events, Language, and Tone), which compiles global news feeds. They cross-referenced data on 9,267 instances of local political unrest in China from 2014 to 2020 with nearly 3 million government contracts issued between 2013 and 2019. The analysis revealed a systematic increase in purchases of cameras and AI-powered facial recognition systems by local authorities in regions following episodes of public unrest.
The direct impact of this equipment on stifling opposition remains challenging to quantify, though researchers contend that it likely contributes to this end. Their findings indicate that these technologies were deployed in response to protests, and each time, the level of protest subsequently diminished. Additionally, economic development associated with these tools was evident in companies that secured contracts, as they went on to produce nearly 50% more technological tools over the following two years.
In essence, the government’s push toward authoritarianism has provided a significant impetus for the development of technologies, particularly in the realm of AI. The study suggests that such growth would have been unlikely without this dynamic. Researchers are now delving deeper into their findings, especially regarding the exportation of these advanced technologies.
They posit that the global marketing of facial recognition tools indicates that government repression could become a widespread global phenomenon. Indeed, various nations, including France, have shown a strong interest in these technologies, as evidenced by legislation like the “2024 Olympic Games Law,” which includes experimentation with facial recognition-equipped cameras at major public events. Recently, the French Senate also voted in favor of using AI-powered systems for real-time surveillance in public spaces, a move contrary to the European Union’s emphasis on strict regulation of AI to ensure the careful monitoring of populations.