This paper critically reviews some measurements of labour market exposure to artificial intelligence (AI) that have been proposed in the literature. Specifically, a methodology that measures the degree of exposure of the activities performed in each occupation is applied to the case of Italian workers, distinguishing between the complementarity and substitutability of AI with these activities. Descriptive evidence on the mobility of workers between occupations with different levels of AI exposure is also provided.
The occupations most exposed to AI are mainly in the service sectors, which employ a significant proportion of women and skilled workers. The majority of highly exposed workers (5 million out of 9) could benefit from AI, as it complements their tasks. Mobility between occupations in different classes of AI exposure is currently low, and workers leaving highly exposed and substitutable occupations mainly move to less exposed occupations, where wages tend to be lower.
