Vasilieva Inna V.

Background. Artificial intelligence (AI) technologies are advancing at an incredible pace, becoming accessible to users. Many on-line services generate photographs, opening opportunities for the creation of deepfake images. Understanding the factors that enhance an individual’s ability to recognize AI-generated images is important for ensuring information security. It is equally important to explore the psychological characteristics that enable individuals to adapt to new technologies.

Objectives. To investigate the relationship between communicative personality characteristics and the ability to distinguish AI-generated images from real photographs.

Study Participants. The study involved 96 respondents aged 18 to 58 years, students and specialists of various profiles (m — 33, f — 63, Mage — 22,27, SD — 7,87).

Methods. The study employed two questionnaires: the Emotional Empathy Test (Y. Orlov, Y. Emelyanov) and the Sociability Assessment Test (V. Ryakhovsky). Stimuli included real photographs and AI-generated images (Kandinsky 2.2, the resource thispersondoesnotexist.com, based on the StyleGAN model).

Results. The study revealed no significant relationship between empathy and sociability and the variable “generation recognition efficiency”. A statistically significant relationship was found between the success in identifying photographs and the variable “experience in interacting with AI”.

Conclusions. Levels of empathy and sociability are not related to the ability to recognize AI-generated images. However, the individuals previously interacting with AI recognize such content more effectively than those having no similar experience. These individuals are better equipped to detect the use of AI by malicious actors. The results obtained in the study contribute to research of AI perception and interaction.