Artificial intelligence reveals propensity for corruption

Researchers at the Vienna University of Economics and Business have developed a model based on artificial intelligence who can predict the inclination of civil servants to corruption based on their personal values ​​and attitudes, announced the Office of the City of Vienna in Belgrade.

Attitudes towards democracy, competition and leadership are more reliable indicators of propensity for corruption than education level or income, according to the study, conducted by Moritz Schmidt and Jürgen Willems.

The research included data on 18.000 civil servants from 90 countries, and was based on an analysis of their views on issues such as the justification of bribery, tax evasion and illegal exercise of the right to state benefits.

According to the researchers, it is difficult to directly measure corrupt behavior at the individual level, which is why the respondents' attitudes towards various forms of illegal behavior were used as an indicator.

In addition to classical statistical methods, the research team applied multiple models of explanatory machine learning, which enabled a more accurate prediction and comparison of the importance of more than 100 possible factors that influence the tendency to corruption.

The results of the research indicate that a strong commitment to democratic values ​​is associated with a lower tendency towards corruption, the announcement states.

Source: FoNet

Real journalism costs money, and we will not be bought by tycoons and corporations. Support us with a one-time or monthly donation. The time for it is now!