Editorial: Artificial intelligence and the future of work: humans in control

Abstract

Latest developments around artificial intelligence (AI) have triggered excitement about the potential to replace and complement human activities while also raising concerns about possible risks to society. Dramatic effects are specifically being felt in the world of work, including jobs, wages and working conditions but also recruitment, performance monitoring, and dismissal. So far, research in this area has focused predominantly on the potential of AI for job gains and losses. Other aspects of its transformative dynamics have received less attention, however. In particular, the impact of AI on job quality, average hours worked, mobility or labour relations between employers and workers are often overlooked. Moreover, society-wide effects triggered by AI, including its rising environmental burden, need to be reassessed. To address these issues, this Research Topic includes nine exciting contributions that shed light on a broader range of issues that AI technologies might bring to the world of work.To set the stage for the overall effect of AI on employment in 23 OECD countries, in our special edition, Georgieff and Hyee present research using an adapted AI occupational impact measure. The authors do not find that AI exposure affects employment growth in their sample. However, occupations where computer use is high see faster employment growth when exposed to AI. In contrast, occupations with low computer use see a decline in average hours work (yet not in employment) when exposed to AI, suggesting a distributional impact of AI rather than one on the overall number of jobs.Whether digital technological technologies improve or worsen wages for employees remains a hotly debated topic. Fossen, Samaan and Sorgner argue that it depends on the specific application considered. Whereas software and industrial robots seem to be associated with wage decreases, suggesting job displacement, innovations in AI are associated with wage increases, pointing towards positive productivity effects, at least as far as the labour market in the United States is concerned.How can the income and wealth disparities that are brought about by AI be addressed? Merola looks at the various proposals that have been brought forward in recent years to address the differential effects of AI on labour markets. She discusses pros and cons of various proposals, including a robot tax, digital taxation, share price taxation or -alternatively -wage subsidies for low-income earners and assesses their potential impact on employment growth, inequality and innovation.Besides the impact of AI on the number of jobs or their distribution, AI will also affect working conditions for those in employment. Using a representative business survey for Germany, Warning, Weber and Püffel demonstrate how occupations with a high share of routine cognitive activities exposed to AI are associated with higher demand for flexibility, including employee self-organisation and time management. Moreover, such worsening of working conditions predominantly affects older workers and women in the labour market.Concerns about implications of AI for occupational health and safety (OSH) abound. Niehaus, Hartwig, Rosen and Wischniewski report results from a large-scale study of German workers on the impact of AI on job autonomy and psychological occupational stress. The authors highlight that AI is often being used to increase autonomy of supervisory functions while lowering control for job execution. This is likely to increase work-related stress over and above possible concerns for job or earnings loss.AI is also a tool that can be used by HR managers to improve functional mobility and ultimately employees' job satisfaction, given that internal mobility risks deteriorating job satisfaction if it increases stress to the detriment of personal life. Bossi et al. analyse various approaches using AI to help better manage internal mobility schemes with a view to improving future job satisfaction. The auth