Porozumění vnímání dopadů umělé inteligence na soukromí v digitálním světě vysokoškolsky vzdělanou veřejností: Výzkum pomocí polostrukturovaných rozhovorů

The aim of this thesis is to determine how university-educated members of the public perceive the risks associated with data privacy in the digital world in the context of artificial intelligence, as identified in academic literature. Methodologically, the thesis combines structured literature research (38 identified studies) with semi-structured interviews with 16 respondents, which lasted 575 minutes. This is followed by a thematic analysis, which generated 1,185 coded text segments. These segments gave rise to seven thematic areas: perceptions of data collection and use in the context of artificial intelligence, artificial intelligence as an amplifier of privacy risks, personal strategies and approaches to privacy protection, commercialisation and monetisation of data, regulation and international legislative differences, awareness, digital literacy and recommendations. The study finds that respondents are aware of the ‘data for services’ business model. This model is perceived as an ubiquitous, inevitable part of the digital world. Respondents view the model pragmatically, often with resignation (11 respondents expressed a certain degree of perceived inevitability). The main deciding factor for using services that implement this model is the level of perceived benefit. This perceived benefit can then be modulated by other factors, such as social pressures from the environment or, for example, distrust of data anonymisation, which was reported by 9 respondents. Respondents also express dissatisfaction with the current level of transparency and communication regarding data flows in the services they use. Another finding is that respondents do not perceive international legislative differences as a major problem that strongly affects data privacy. However, China is an exception. Respondents view this international region with a certain degree of scepticism regarding the security of local applications. The attitude of respondents can therefore be defined as rather selective. Respondents are most knowledgeable about their local legislation (GDPR), while their knowledge of other non-European legislative frameworks is only superficial. The exception is respondents with legal education, who demonstrate a higher overall level of knowledge. In terms of specific recommendations for improving the situation, the report recommends: improving the clarity and transparency of existing and upcoming legislation on personal data protection and data flow regulation; (2) increasing pressure from regulatory authorities on digital service providers to openly communicate and monitor user data flows; (3) systematically apply the principles of privacy-by-design, regular external audits, precise risk categorisation and also allow opt-outs in cases where users do not want to participate in data transactions. These measures will lead to a reduction in information asymmetry and strengthen the decision-making power of service users.