Abstract
Corruption crime is not limited to the deviance of one actor, but is a crime with destructive power that can destroy the social structure. Although the state and society are making great efforts to reduce corruption crime, it is not easily detected or completely eradicated due to the confidentiality and customary nature of corruption crime. Therefore, it is necessary to explore whether there is an effective way to reduce corruption by introducing an AI crime prediction system. This is because many countries are already seeing a positive effect of reducing the crime rate by introducing AI crime prediction systems. However, it is necessary to examine whether the AI crime prediction system can be applied to corruption crimes and what problems there are if it is applied. Furthermore, we intend to explore the improvement measures for applying the AI crime prediction system to corruption crimes. AI crime prediction systems that handle personal information can infringe on individual rights. And the algorithm of AI crime prediction system has problems of opacity, inaccuracy, and bias. AI crime prediction systems are becoming a new means of control for citizens, such as ‘Big Brother’. As a solution to bias and inaccuracy, the AI crime prediction system should be descriptive, accurate, and audited by an objective agency. In order to limit the individual control function of the AI crime prediction system, the AI crime prediction system must also apply the principle of proportionality to minimize the infringement of the subject's rights (the principle of minimal infringement). And AI crime prediction systems should be used as a last resort (principle of complementarity). It should be used individually according to the subject's risk of recidivism (principle of individuality). The AI crime prediction system should not be implemented by administrative disposition, but should be imposed by judicial review of the court (principle of judicial control). If the analysis of the AI crime prediction system is applied according to the principle of imposition of such security measures, the individual control function of the AI crime prediction system can be limited as much as possible and the efficiency as a control method for corruption crimes can be increased.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call