What shapes our attitudes towards algorithms in urban governance? The role of perceived friendliness and controllability of the city, and human-algorithm cooperation

Today, artificial intelligence (AI) algorithms are increasingly often implemented in urban management; however, so far, there has been no research on how their use can affect the perception of the city. Is a city employing algorithmic decision-making perceived as friendly and controllable? Will algorithms assisting city officials be more accepted than those making a decision without human supervision? And, is a city in which algorithms make decisions with city officials perceived as more friendly and controllable than a city in which algorithms make decisions unadvised? To address these questions, we have conducted two studies in Poland, representative for the population's age and gender structure. The first study results (N = 753) showed that individuals who perceived cities using algorithms as resident-unfriendly, were also more reluctant to accept algorithmic governance. The perceived controllability of the city was not related to the acceptance of algorithms. The second, experimental study (N = 1047) demonstrated that the acceptance of algorithms increased when they were presented as collaborating with humans (vs. making decisions unadvised). Cooperation between algorithms and humans also increased the perceived friendliness and controllability of the city. The studies provide evidence that imple-menting algorithms into urban governance can lead to a sense of alienation and lower the city's friendliness; however, collaboration between humans and algorithms may diminish this feeling and increase the overall acceptance of algorithmic decision making in the city.

Automated summary

Currently, there is a lack of research on how the use of artificial intelligence algorithms can impact the perception of cities. Two studies conducted in Poland found that individuals who viewed cities using algorithms as unfriendly were also less likely to accept algorithmic governance. However, the perceived controllability of the city was not related to the acceptance of algorithms. A second experimental study showed that the acceptance of algorithms increased when they were presented as collaborating with humans, leading to an increase in the perceived friendliness and controllability of the city.