AI in the Job Market Could Increase Inequalities Instead of Reducing Them

According to a new study by the Institute for Advanced Studies (IHS), artificial intelligence (AI) does not have a significant impact on income and job security. While highly skilled employees see slight income increases, the risk of layoffs and income loss rises for less skilled workers. AI could therefore even exacerbate existing inequalities, according to the authors.

AI Skills Hardly Demanded in the Labor Market

For the analysis, around eight million online job advertisements in Germany between 2017 and 2023 were evaluated and linked with employment data from two million workers. The proportion of job postings requiring AI skills is only around 1 to 1.5 percent. The study focused primarily on early AI technologies such as machine learning, where computers learn to recognize patterns and make predictions.

According to the study, such early AI technologies primarily benefit highly skilled professionals. Their income increased by an average of 403 euros per year. "AI opens up new areas of work, especially in knowledge-intensive professions - and these new fields of activity continue to be shaped by humans," explained IHS labor market researcher Eduard Storm.

IHS Doubts Job Losses Due to AI

Overall, AI creates more jobs than it destroys. "I do not believe that there will be widespread job losses, at least in terms of mass unemployment," said Storm during a presentation of the study. New jobs are primarily emerging in analytical professions in the IT sector as well as in interactive professions, such as consulting or sales. There, AI can take over administrative tasks and thus create more space for activities like customer acquisition.

Income Decline for Less Skilled Individuals

For less skilled employees, however, the study shows potentially burdensome trends. For workers with the lowest incomes, AI leads to a reduction in annual working days by eight days and an income decline of 3.9 percent. According to the study authors, these results challenge more optimistic assumptions that AI could reduce social inequality.

(APA/Red)

This article has been automatically translated, read the original article here.