Can AI Transform Music Production? - Intervention Analysis for Short Time Series on the Global Supply of Music Recordings.

The introduction of artificial intelligence (AI) technology to the public has sparked a widespread debate regarding its implications for economy and society, including the cultural and creative industries, within which music. While the debate often focuses on AI's potential to automate certain parts of creative labour or the potential of AI to become a creator itself, empirical evidence about the real-world impact in the music industry remains scarce. This thesis addresses a critical gap in the literature by analysing how the advent of generative AI tools has affected the global supply of music recordings. ?The study applies an intervention analysis framework, similar to methodological approach by Handke (2012) for digital copying, conceptualising it as a natural experiment. Using global release data from the MusicBrainz database, the thesis covers the years from 2000 to 2024 and tests whether the supply of new music underwent a significant structural change after 2016, which is a period identified as the first year in which music AI production programs were introduced, the beginning of a widespread AI adoption in music creation. ?To assess the effect of AI on the global supply of music recordings, the analysis applies the C statistic, first developed by Young (1941) and applied by Tryon (1982). Linear regression and first differencing are used to remove any deterministic patterns from the ore-AI period (2000-2015) in order to correctly assess the C statistic. The Z statistic derived from the C statistic and applied to the full time series (2000-2024) reveal statistically significant structural change for the period after introduction of AI, which indicates that AI introduction coincides with a disruption in supply dynamics. ?Further analysis at the genre level is conducted between technology-dependent genres (pop, dance, techno) and traditionally acoustic genres (jazz, classical). The results of the genre-specific analysis show that all genres experienced statistically significant change, the effect was most pronounced in pop and jazz genres, rejecting the alternative hypothesis that AI integration impacts pop, dance and techno genres more than jazz and classical. Overall, the findings suggest that the introduction of AI into music production has not only sustained the existing growth trends but altered the underlying structure of supply of music recordings. The results contribute to the field of cultural economics by providing empirical evidence of how digital technologies might possibly have a significant impact on the supply of music recordings and a music sector overall. Researchers are invited to expand this work by exploring the potential for the transformative ways that AI may bring to the cultural and creative sectors in Europe and globally.