The application of artificial intelligence in people's daily life has become inevitable, especially in the business environment. However, most organizations acknowledge the fact that the use of AI Systems is necessary, but guidelines are also necessary to adapt to the accelerated expansion of artificial intelligence. Security systems represent a particularly sensitive category when it comes to the application of artificial intelligence, given the huge unknowns about the risks that these systems carry with them. A substantive framework for risk assessment in this area is imposed as necessary because the existing risk assessment methods generally do not recognize AI risks, which prevents their comprehensive overview. This paper provides a comprehensive conceptual content of the risk assessment of the application of AI in security systems, which includes risks during the entire life cycle of the AI system, which enables the identification, analysis, assessment and treatment of recognized risks and the monitoring and control of their impact on the system.
Mladenović et al. (Thu,) studied this question.