Poptávka po dovednostech v oblasti umělé inteligence v inzerátech IT pozic

Abstract

This master's thesis examines the demand for artificial intelligence skills in IT job postings through a comparative analysis of three economically and institutionally distinct markets: the United States, Germany, and India. The theoretical part focuses on the foundations of human capital theory, the conceptualisation of artificial intelligence as a general-purpose technology, and its impact on the task composition of employment. The empirical part draws on a sample of 38,436 job postings obtained from Glassdoor, processed through a combination of deterministic skill extraction and classification by a large language model. This approach makes it possible to distinguish positions without any AI requirement, positions that only integrate AI into existing workflows, and positions focused on the development of AI itself. Binary and multinomial logistic regressions are used to identify the skill profiles and occupational groups associated with AI requirements, and the wage premium for AI skills is subsequently quantified through an OLS regression based on the Mincer wage equation. The results show that the adoption of AI skills differs across the three markets and that AI requirements are strongly tied to a modern technological profile of the position. The thesis contributes to the discussion on the transformation of human capital in the context of technological change and offers a methodological framework for analysing AI demand based on granular data from job postings.