Generativní umělá inteligence pro syntézu obrazu

This bachelor's thesis is devoted to the identification and analysis of potential risks brought by generative artificial intelligence in the field of image synthesis. As part of the theoretical part, a literature review was carried out, on the basis of which the key terms associated with generative artificial intelligence and its components enabling image synthesis were defined. The work also presents possible risks and provides users with a basic methodology for their mitigation. In the practical part, quantitative research was carried out through a questionnaire survey with the aim of answering the research questions. A total of 183 respondents took part in the questionnaire and were also tested on their ability to distinguish synthetic images from real photographs. The research results showed that almost half of the respondents have personal experience with generative artificial intelligence for image synthesis and are aware of the risks associated with it. Respondents also confirmed that this technology is being misused for malicious purposes, including manipulation of public opinion. On average, research participants were able to identify approximately half of the synthetic images. In the future, the increasingly convincing outputs of generative artificial intelligence will represent a considerable challenge. Therefore, it is necessary to educate the general public and alert them to possible threats