Return of the Solow-paradox in AI?: AI-adoption and firm productivity

AI-related technologies have become ubiquitous in many contexts within the business world. However, to date, empirical accounts of the productivity effects of AI adoption are scarce. Using data from Finnish job advertisements posted between 2013 and 2019, we identify job advertisements referring to AI-related skills. Matching this data to productivity data from ORBIS, we then estimate the productivity effects of AI-related activities in our sample. Our results indicate that AI-adoption increases productivity—with three important qualifications. Firstly, effects are only observable for large firms with more than 499 employees. Secondly, there is evidence that early adopters did not experience productivity increases, which could be attributed to the technological immaturity of earlier generations of AI. Thirdly, we find evidence showing a minimum delay of three years between AI adoption and ensuing productivity effects (i.e., the investment delay effect). We argue that our findings related to the impact of technological immaturity and the investment delay effect may help explain the return of the so-called AI-related Solow-paradox: AI is everywhere except in productivity statistics.