A recent study states that AI technology—the basis of a "fourth industrial revolution"—may allow frontier innovation and autocracy to coexist. AI's predictive power may help autocrats govern society.

Government AI purchases may also spur innovation, like dual-use technologies. Because government data can be used in AI prediction algorithms and shared across purposes, autocracies' data collecting and processing for political control may inspire commercial AI innovation outside government applications. These theories suggest a mutually reinforcing connection in which governments buy AI to gain political power and the technology innovates.

According to the paper, "AI innovation entrenches the regime, and the regime's investment in AI for political control stimulates further frontier innovation." Scholars call this situation an "AI-tocracy," characterising a cycle in which further deployment of AI-driven technology quells opposition while increasing the country's innovative capacity.

GDELT Project

The researchers used a variety of evidence over the last decade to perform the study. They used data from the Global Database of Events, Language, and Tone (GDELT) Project, which archives news feeds worldwide, to categorise occurrences of political upheaval. Between 2014 and 2020, the team discovered 9,267 incidences of discontent.

The researchers then reviewed data from a database on around 3 million procurement contracts. They discovered that local governments' purchase of facial-recognition AI services and related public security technologies — high-resolution video cameras — increased dramatically in the quarter after an outbreak of public unrest in that area.

Facial recognition technology

The researchers believe it did, albeit they note in the study that they "cannot directly estimate the effect" of the technology on political turmoil. However, as one approach to answering that question, they investigated the association between weather and political upheaval. In doing so, the researchers also considered whether higher relative wealth levels in particular places would have resulted in increased investments in AI-driven technology regardless of protest patterns. Nonetheless, the experts came to the same conclusion: facial recognition technology was implemented in response to previous protests, decreasing further protest levels.

Evaluation

Lastly, the research team looked at how the increased demand for AI affected the technology sector. They found that the government's increased use of facial-recognition tools. For example, companies that get government contracts to buy facial-recognition technology make about 49% more software goods two years after getting the contract than before. The result also has implications for more significant questions about how governments work and the economy grows. 

Conclusion

Many academic studies show that institutions that give people rights lead to more economic growth over time by making new technologies more accessible. Still, looking at the benefits of using AI shows one-way authoritarian governments can grow faster than they would have otherwise.

Image source: Unsplash