Digitale Arbeitswelt – Chancen und Herausforderungen für Beschäftigte und Arbeitsmarkt
Der digitale Wandel der Arbeitswelt gilt als eine der großen Herausforderungen für Wirtschaft und Gesellschaft. Wie arbeiten wir in Zukunft? Welche Auswirkungen hat die Digitalisierung und die Nutzung Künstlicher Intelligenz auf Beschäftigung und Arbeitsmarkt? Welche Qualifikationen werden künftig benötigt? Wie verändern sich Tätigkeiten und Berufe? Welche arbeits- und sozialrechtlichen Konsequenzen ergeben sich daraus?
Dieses Themendossier dokumentiert Forschungsergebnisse zum Thema in den verschiedenen Wirtschaftsbereichen und Regionen.
Im Filter „Autorenschaft“ können Sie auf IAB-(Mit-)Autorenschaft eingrenzen.
- Gesamtbetrachtungen/Positionen
- Arbeitsformen, Arbeitszeit und Gesundheit
- Qualifikationsanforderungen und Berufe
- Arbeitsplatz- und Beschäftigungseffekte
- Wirtschaftsbereiche
- Arbeits- und sozialrechtliche Aspekte / digitale soziale Sicherung
- Deutschland
- Andere Länder/ internationaler Vergleich
- Besondere Personengruppen
-
Literaturhinweis
Strukturwandel in Thüringen: Digitalisierung. Mit einer Neuschätzung der Substituierbarkeitspotenziale (2026)
Kropp, Per; Fritzsche, Birgit; Theuer, Stefan;Zitatform
Kropp, Per, Stefan Theuer & Birgit Fritzsche (2026): Strukturwandel in Thüringen: Digitalisierung. Mit einer Neuschätzung der Substituierbarkeitspotenziale. (IAB-Regional. Berichte und Analysen aus dem Regionalen Forschungsnetz. IAB Sachsen-Anhalt-Thüringen 02/2026), Nürnberg, 42 S. DOI:10.48720/IAB.RESAT.2602
Abstract
"Die Arbeitswelt verändert sich in rasant. Technische Entwicklungen bei Software, Computer oder computergesteuerten Maschinen schaffen immer neue Anwendungsmöglichkeiten. Bislang waren insbesondere Routinetätigkeiten z. B. bei Helfertätigkeiten automatisierbar. Nun sind durch produktiv nutzbare KI-Technologie auch zunehmend nicht-routine-Tätigkeiten von Spezialisten und Experten betroffen. Im Vergleich zu Deutschland hatte Thüringen häufiger Berufe mit einem hohen Substituierbarkeitspotenzial. Das durchschnittliche Substituierbarkeitspotenzial über alle Berufe stieg bis 2016 rasant an, seitdem jedoch im geringeren Ausmaß. Sicherheitsberufe hatten 2019 mit rund 21 Prozentpunkten einen sehr hohen Anstieg, so wie zuletzt die IT- und naturwissenschaftlichen Dienstleistungsberufe mit rund 20 Prozentpunkten. Für diese Veränderungen konnten drei Faktoren identifiziert werden: die Veränderung der Substituierbarkeitspotenziale einzelner Berufe, innerberufliche Veränderungen wie bei den Kerntätigkeiten und der berufliche Strukturwandel. Für Männer und Frauen sind die Substituierbarkeitspotenziale seit 2013 ähnlich gestiegen, bei Männern allerdings von einem höheren Anfangsniveau. KI ersetzt dabei eher Tätigkeiten, die mehrheitlich von Frauen erledigt werden. Zwischen den verschiedenen Alterskohorten gibt es insgesamt kaum Unterschiede. Lediglich in zwei Berufssegmenten haben jüngere Beschäftigte ein höheres Substitutionspotenzial – bei Verkehrs- und Logistikberufen sowie bei den IT- und naturwissenschaftlichen Dienstleistungsberufen. Auch wenn sich einige Berufe durch Digitalisierung stark verändern führt das kaum zu Beschäftigungsverlusten. Für solche Berufe und für Regionen, in denen sie vermehrt vorkommen, können jedoch höhere Weiterbildungsbedarfe vermutet werden." (Autorenreferat, IAB-Doku)
Weiterführende Informationen
Online-Anhänge zu Substituierbarkeitspotenzialen in Thüringen (nicht barrierefrei) -
Literaturhinweis
Gestaltungsprinzipien individueller Beruflichkeit am Beispiel von Crowdworker:innen: Theoretische Ansätze und empirische Ergebnisse zum Zusammenhang von Arbeit, Beruf und Subjektivierung (2026)
Külpmann, Inga;Zitatform
Külpmann, Inga (2026): Gestaltungsprinzipien individueller Beruflichkeit am Beispiel von Crowdworker:innen. Theoretische Ansätze und empirische Ergebnisse zum Zusammenhang von Arbeit, Beruf und Subjektivierung. (Berufsbildung, Arbeit und Innovation. Dissertationen, Habilitationen 92), Bielefeld: wbv, 352 S. DOI:10.3278/9783763979080
Abstract
"Die Studie untersucht, wie Crowdworker:innen in Plattformarbeit individuelle Beruflichkeit entwickeln und wie Arbeitserfahrungen Subjektivierung prägen. Im Mittelpunkt steht die Frage, welche Gestaltungsprinzipien Lern- und Entwicklungspotenziale in Crowdwork unterstützen. Aus berufs- und wirtschaftspädagogischer sowie arbeitssoziologischer Perspektive verbindet die Publikation theoretische Ansätze zu Arbeit, Beruf und Subjektivierung mit empirischen Ergebnissen aus Untersuchungen von Crowdwork-Plattformen in Deutschland. Die Ergebnisse werden im Kontext des Forschungsprojekts CKoBeLeP (2021-2024) eingeordnet und auf Implikationen für lernförderliche Plattformgestaltung bezogen. Geeignet für Studierende und Forschende der Berufs- und Wirtschaftspädagogik, Arbeitssoziologie und Digitalisierungsforschung sowie für Akteur:innen in Arbeitsgestaltung, Weiterbildung und Plattformregulierung." (Verlagsangaben, IAB-Doku)
-
Literaturhinweis
Automation and the risk of labor market exclusion across Europe (2026)
Zitatform
Lamperti, Fabio & Davide Castellani (2026): Automation and the risk of labor market exclusion across Europe. In: Structural Change and Economic Dynamics, Jg. 77, S. 62-76. DOI:10.1016/j.strueco.2025.12.014
Abstract
"Labor market exclusion represents a major concern in several European economies, particularly affecting highly exposed demographic groups. This paper examines the potential effect of automation technologies on the risk of being locked into protracted unemployment or inactivity, using Labour Force Survey data for the European Union 27 countries and the United Kingdom, between 2009 and 2019. Our study employs repeated cross-sections of individual-level data to compute probabilities of exclusion outcomes due to automation adoption, controlling for several individual, macroeconomic, and region-specific characteristics, and for potential selection mechanisms. Findings highlight that, on average, the adoption of new automation technologies is associated with a higher probability of being inactive. This is consistent with the view that automation may exacerbate job insecurity, psychological discouragement, and detachment from job-seeking. This relationship is heterogeneous across demographic groups, with younger individuals being relatively more affected." (Author's abstract, IAB-Doku, © 2025 The Authors. Published by Elsevier B.V.) ((en))
-
Literaturhinweis
Automation, Trade Unions and Atypical Employment (2026)
Zitatform
Lewandowski, Piotr & Wojciech Szymczak (2026): Automation, Trade Unions and Atypical Employment. In: Industrial Relations, Jg. 65, H. 3, S. 378-396. DOI:10.1111/irel.70017
Abstract
"We study the effect of automation technologies—industrial robots, software and databases—on the incidence of involuntary atypical employment in 13 EU countries between 2006 and 2018. Robots do not affect the total employment rate but significantly increase the involuntary atypical employment share, mainly through fixed-term work. Software and databases increase total employment and are neutral for atypical employment. Higher trade union density mitigates the robots' impact on atypical employment, while employment protection legislation plays no role. Using historical decompositions, we attribute 1–2 percentage points of a 15% average atypical employment share in our sample to automation." (Author's abstract, IAB-Doku, Published by arrangement with John Wiley & Sons) ((en))
-
Literaturhinweis
Economics of Human and AI Collaboration: When is Partial Automation More Attractive than Full Automation? (2026)
Li, Wensu; Lyu, Harry; Goehring, Brian C.; Aboutorabi, Atin; Qian, Kaizhi; Thompson, Neil; Fleming, Martin;Zitatform
Li, Wensu, Atin Aboutorabi, Harry Lyu, Kaizhi Qian, Martin Fleming, Brian C. Goehring & Neil Thompson (2026): Economics of Human and AI Collaboration: When is Partial Automation More Attractive than Full Automation? (arXiv papers 2603.29121), 57 S.
Abstract
"This paper develops a unified framework for evaluating the optimal degree of task automation. Moving beyond binary automate-or-not assessments, we model automation intensity as a continuous choice in which firms minimize costs by selecting an AI accuracy level, from no automation through partial human-AI collaboration to full automation. On the supply side, we estimate an AI production function via scaling-law experiments linking performance to data, compute, and model size. Because AI systems exhibit predictable but diminishing returns to these inputs, the cost of higher accuracy is convex: good performance may be inexpensive, but near-perfect accuracy is disproportionately costly. Full automation is therefore often not cost-minimizing; partial automation, where firms retain human workers for residual tasks, frequently emerges as the equilibrium. On the demand side, we introduce an entropy-based measure of task complexity that maps model accuracy into a labor substitution ratio, quantifying human labor displacement at each accuracy level. We calibrate the framework with O*NET task data, a survey of 3,778 domain experts, and GPT-4o-derived task decompositions, implementing it in computer vision. Task complexity shapes substitution: low-complexity tasks see high substitution, while high-complexity tasks favor limited partial automation. Scale of deployment is a key determinant: AI-as-a-Service and AI agents spread fixed costs across users, sharply expanding economically viable tasks. At the firm level, cost-effective automation captures approximately 11% of computer-vision-exposed labor compensation; under economy-wide deployment, this share rises sharply. Since other AI systems exhibit similar scaling-law economics, our mechanisms extend beyond computer vision, reinforcing that partial automation is often the economically rational long-run outcome, not merely a transitional phase." (Author's abstract, IAB-Doku) ((en))
-
Literaturhinweis
Skill-biased technological change in the age of AI: a theoretical analysis of automation and inequality (2026)
Li, Te; Nichols, Hadrian;Zitatform
Li, Te & Hadrian Nichols (2026): Skill-biased technological change in the age of AI: a theoretical analysis of automation and inequality. In: Economics of Innovation and New Technology, S. 1-29. DOI:10.1080/10438599.2026.2649376
Abstract
"This paper develops a general equilibrium model to analyze how artificial intelligence (AI)–driven automation reshapes productivity, labor markets, and income distribution. The model features heterogeneous workers, endogenous automation decisions, and irreversible skill investment choices, allowing a unified examination of displacement, complementarity, and skill-supply responses. Automation substitutes for labor in routine tasks, reducing demand and wages for low-skilled workers, while simultaneously enhancing productivity in complex tasks where AI complements high-skilled labor. As a result, aggregate output and total welfare increase, but income inequality widens and the skill premium rises. A key finding is that technological progress is not Pareto improving: due to heterogeneous and irreversible skill investment costs, workers below a critical ability threshold experience absolute welfare losses despite overall economic growth. This generates a ‘growth paradox’ in which productivity gains coexist with immiseration for vulnerable groups. The paper further evaluates three policy instruments – redistributive taxation, education subsidies, and technology policy – and derives conditions under which each improves social welfare. Education subsidies emerge as the most effective tool for mitigating inequality while preserving efficiency, whereas excessive automation may arise when private incentives diverge from social optima." (Author's abstract, IAB-Doku) ((en))
-
Literaturhinweis
AI adoption among German firms (2026)
Zitatform
Licht, Thomas & Klaus Wohlrabe (2026): AI adoption among German firms. In: Structural Change and Economic Dynamics, S. 1-5. DOI:10.1080/13504851.2026.2669986
Abstract
"This paper examines the adoption of Artificial Intelligence (AI) among German firms using firm-level data from the ifo Business Survey. Between 2023 and 2024, AI usage more than doubled from 13.3% to 27%, with substantial variation across sectors and firm sizes. Large firms in manufacturing and services lead adoption, while SMEs and construction companies lag behind. Using linear probability models with fixed effects for industry and firm size, we find that managerial characteristics – particularly risk tolerance and patience – significantly influence adoption decisions. Firms led by more risk-tolerant and patient managers are more likely to implement AI. Most firms expect notable productivity gains from AI – approximately 8% overfive years – though expectations vary by sector. Our findings providenew firm-level evidence by linking validated measures of managerial preferences to AI adoption decisions, highlighting both the economic potential and the structural barriers to broader diffusion." (Author's abstract, IAB-Doku) ((en))
-
Literaturhinweis
Good Jobs or Bad Jobs? Immigrant Workers in the Gig Economy (2026)
Zitatform
Liu, Cathy Yang & Rory Renzy (2026): Good Jobs or Bad Jobs? Immigrant Workers in the Gig Economy. In: International migration review, Jg. 60, H. 1, S. 114-138. DOI:10.1177/01979183241309585
Abstract
"New work arrangements enabled by online platforms, or gig work, saw substantive growth during the COVID-19 pandemic. Various estimates have suggested the wide participation of workers in the gig economy, with minority and immigrant workers well represented. The quality of work is a multi-dimensional concept that goes beyond earnings. One framework of good jobs and bad jobs centers on control over work schedule, content and duration, stability, safety, benefits and insurance, as well as career advancement opportunities. Using a newly released national survey focused on entrepreneurs and workers in the United States, we find that about 18.5 percent immigrant workers and 21.1 percent native-born workers participated in the gig economy as their primary or secondary job. In terms of job quality, immigrant gig workers work shorter hours and have significantly less fringe benefits than non-gig workers as well as U.S.-born gig workers, reflecting a double disadvantage. However, they tend to have higher entrepreneurial aspirations, suggesting the transient nature of gig arrangements and potential for career advancements. This paper provides a comprehensive analysis of the characteristics and implication of immigrants’ engagement with the gig economy and offers policy and theoretical discussions." (Author's abstract, IAB-Doku) ((en))
-
Literaturhinweis
The Enshittification of Work: Platform Decay and Labour Conditions in the Gig Economy (2026)
Zitatform
Maffie, Michael David & Hector Hurtado (2026): The Enshittification of Work: Platform Decay and Labour Conditions in the Gig Economy. In: BJIR, Jg. 64, H. 1, S. 5-20. DOI:10.1111/bjir.70004
Abstract
"This study investigates the mechanisms by which gig platforms degrade labor conditions over time, building on the concept of platform decay, or ‘enshittification’, initially developed in the context of social media platforms. In this article, we draw on 30 interviews with long-term gig workers in the ride-hail and grocery delivery sectors, offering insights into how these companies shift from offering attractive working conditions to exploiting labor as these services develop market power via network effects. We identify three mechanisms through which gig companies claw back value from workers over time: burden shifting (transferring operational costs to workers), feature addition and alteration (increasing the demands on workers), and market manipulation (reducing worker bargaining power). We then explore how workers respond to platform decay, finding that workers adopt three responses: effort recalibration , multi-homing and navigating the changing conditions through what we term toxic resilience . This study contributes to the gig work literature by developing a framework to explain how working conditions in the gig economy improve or degrade over time. In doing so, this article provides a framework for organizing the growing constellation of labour research on gig workers." (Author's abstract, IAB-Doku, Published by arrangement with John Wiley & Sons) ((en))
-
Literaturhinweis
Monitor Digitale Arbeitsgesellschaft – Kurzbericht der dritten Befragungswelle. Nach dem KI-Aufmerksamkeitsanstieg: Zwischen Stabilisierung, Skepsis und Regulierungsanforderungen (2026)
Zitatform
Marcinkowski, Frank, Birte Keller, Marco Lünich & Florian Golo Flaßhoff (2026): Monitor Digitale Arbeitsgesellschaft – Kurzbericht der dritten Befragungswelle. Nach dem KI-Aufmerksamkeitsanstieg: Zwischen Stabilisierung, Skepsis und Regulierungsanforderungen. (SocArXiv papers), 11 S. DOI:10.31235/osf.io/zcxje_v1
Abstract
"Der Kurzbericht präsentiert zentrale Ergebnisse der dritten von vier repräsentativen Befragungswellen der Studie Monitor Digitale Arbeitsgesellschaft, die im Rahmen des Forschungsprojekts Meinungsmonitor Künstliche Intelligenz 3.0 durchgeführt wurde. Grundlage sind die Angaben von 1.647 Befragten (928 Erwerbstätige, 719 Nichterwerbstätige) im Januar 2026. Die dritte Befragungswelle zeigt, dass das im Verlauf des Jahres 2025 gestiegene Interesse an KI zu Beginn des Jahres 2026 wieder leicht zurückgeht, während die subjektive Einschätzung des eigenen Wissens auf dem erreichten Niveau stabil bleibt. Gleichzeitig nehmen Vorbehalte gegenüber bestimmten gesellschaftlichen Einsatzfeldern zu, insbesondere in Bereichen mit Bezug zu staatlichen Sicherheitsaufgaben und politischen Entscheidungen. In der Arbeitswelt bleibt die Bewertung der Auswirkungen von KI auf den eigenen Beruf insgesamt stabil und leicht nutzenorientiert. Die berufliche Nutzung von KI stagniert zuletzt weitgehend. Viele Beschäftigte berichten weiterhin keine grundlegenden Veränderungen ihrer Tätigkeitsanforderungen und nur geringe persönliche Jobunsicherheit, während die Sorge, dass KI menschliche Arbeit allgemein ersetzen oder grundlegend verändern könnte, weiterhin deutlich stärker verbreitet ist als die persönliche Jobunsicherheit. Erstmals wird zudem ein besonderes Schlaglicht auf die Regulierungspräferenzen der Bevölkerung geworfen: Während staatliche Eingriffe bei arbeitsmarktbezogenen Folgen breite Unterstützung finden, werden Eingriffe in wirtschaftliche Freiheiten und steuerpolitische Maßnahmen deutlich ambivalenter bewertet. Insgesamt deutet sich eine stärkere Ausdifferenzierung und Festigung der Einstellungen gegenüber KI an." (Autorenreferat, IAB-Doku)
-
Literaturhinweis
Improving the effects of industrial robot adoption on employment, total factor productivity, and real wages in 52 world economies and OECD members (2026)
Zitatform
Matsuki, Takashi (2026): Improving the effects of industrial robot adoption on employment, total factor productivity, and real wages in 52 world economies and OECD members. In: Review of world economics, Jg. 162, H. 2, S. 417-448. DOI:10.1007/s10290-025-00626-z
Abstract
"This study investigates the effects of industrial robot adoption in the production process on unemployment rate, employment ratio in manufacturing, and total factor productivity (TFP) growth in 52 countries, and real wage growth in 31 and 20 OECD member countries for 2007–2019. The operating stock of robots per employee significantly impacts these variables; robot adoption lowers the unemployment rate and raises TFP and real wage growth. However, it reduces the employment ratio in manufacturing. In addition, the slight but significant positive contribution of robot adoption is observed only in the 90-percentile (top 10-percentile) of the real wage distribution. Interestingly, workers in the bottom and top tails (10- and 90-percentiles) of the wage distribution asymmetrically benefit from robotization. The industry ratio of value-added improves the labor market by reducing the unemployment rate and raising the employment ratio in manufacturing, TFP growth, and real wage growth. The information and communication technology (ICT) development also positively contributes to the employment ratio in Asia’s manufacturing, TFP growth, and real wage growth." (Author's abstract, IAB-Doku) ((en))
-
Literaturhinweis
Artificial Intelligence and Productivity in Europe (2026)
Misch, Florian; Park, Ben; Sher, Galen; Pizzinelli, Carlo;Zitatform
Misch, Florian, Ben Park, Carlo Pizzinelli & Galen Sher (2026): Artificial Intelligence and Productivity in Europe. (CESifo working paper 12401), München, 37 S.
Abstract
"The discussion on Artificial Intelligence (AI) often centers around its impact on productivity, but macroeconomic evidence for Europe remains scarce. Using the Acemoglu (2024) approach we simulate the medium-term impact of AI adoption on total factor productivity for 31 European countries. We compile many scenarios by pooling evidence on which tasks will be automatable in the near term, using reduced-form regressions to predict AI adoption across Europe, and considering relevant regulation that restricts AI use heterogeneously across tasks, occupations and sectors. We find that the medium-term productivity gains for Europe as a whole are likely to be modest, at around 1 percent cumulatively over five years. While economically still moderate, these gains are still larger than estimates by Acemoglu (2024) for the US. They vary widely across scenarios and countries and are substantially larger in countries with higher incomes. Furthermore, we show that national and EU regulations around occupation-level requirements, AI safety, and data privacy combined could reduce Europe's productivity gains by over 30 percent if AI exposure were 50 percent lower in tasks, occupations and sectors affected by regulation." (Author's abstract, IAB-Doku) ((en))
-
Literaturhinweis
Predicting and Preventing Turnover in Industry 4.0: Understanding the Impact of Artificial Intelligence Adoption on Employee Turnover (2026)
Moon, Young-Kook; Mitropoulos, Tanya;Zitatform
Moon, Young-Kook & Tanya Mitropoulos (2026): Predicting and Preventing Turnover in Industry 4.0: Understanding the Impact of Artificial Intelligence Adoption on Employee Turnover. In: Human resource development quarterly, S. 1-11. DOI:10.1002/hrdq.70018
Abstract
"With the increasing adoption of Artificial Intelligence (AI) in the workplace, employees' career paths have become more diverse and less predictable in the era of Industry 4.0. As technological transformations accelerate, employee turnover patterns are also changing, as reflected in the growing prevalence of occupational transitions and large-scale layoffs. These shifts highlight the need for more developmental approaches to understanding and managing retention. To address this issue, the present paper examines how AI implementation influences employee turnover by focusing on both the motives and psychological states underlying the withdrawal process. Drawing on proximal withdrawal states theory, the paper provides a deeper understanding of the dynamic nature of employee withdrawal patterns. This perspective elucidates the withdrawal process shaped by AI adoption, emphasizing that employees' responses to technological disruption are diverse in both psychological and behavioral terms rather than uniform. Furthermore, the paper proposes Human Resource Development (HRD)-based interventions aimed at preventing undesirable turnover through employee education and career development initiatives. By integrating these elements, we introduce an HRD-focused framework for addressing turnover issues associated with AI adoption in a more developmental and sustainable manner. Finally, the paper outlines future research directions for empirically testing the proposed theoretical mechanisms and intervention strategies within the rapidly evolving technological landscape of work." (Author's abstract, IAB-Doku) ((en))
-
Literaturhinweis
The remote reality: unpacking the impact of an exogenous shock on online labor markets (2026)
Zitatform
Mourelatos, Evangelos, Jaakko Simonen & Simo Hosio (2026): The remote reality: unpacking the impact of an exogenous shock on online labor markets. In: Journal of business economics, S. 1-41. DOI:10.1007/s11573-026-01270-1
Abstract
"Following the outbreak of COVID-19 and the widespread shift to remote work, interest has grown in how these changes have affected online labor markets. We study individuals who were active in both the conventional labor market and an online labor market prior to this shock, and compare workers who were shifted from on-site to remote work in their main job with those who remained on-site. To provide insights into how changes in working conditions affect productivity in online labor markets, we utilized a unique dataset obtained from a prominent online labor marketplace. Our dataset contains several measures of worker productivity, which we analyze in conjunction with survey data collected from online workers regarding their experiences with conventional market changes. By combining these two sources of data, we are able to investigate the ways in which the pandemic-induced shift toward remote work has affected the productivity and behavior of online workers. We find that, first, the shift toward remote work during the pandemic has led to changes in the patterns of productivity in online labor markets. The transition toward remote work from home (WFH) decreased online productivity by 18%. Second, our findings suggest that the decrease in work quality among workers transitioning to WFH was primarily due to a “sloppy” work approach. Specifically, while the quantity of online work output may increase for such workers, their overall work quality tended to decline compared to those who continued to work in traditional firm environments during the COVID-19 pandemic. Third, heterogeneity analysis reveals that workers with higher levels of neuroticism tended to exhibit better adaptability to the WFH setup by having higher online productivity. Our analysis reveals several important findings with implications for both workers and policy-makers in the gig economy." (Author's abstract, IAB-Doku) ((en))
-
Literaturhinweis
KI in Betrieben: Mehr Ausbildung – aber Weiterbildung zunehmend für anspruchsvollere Tätigkeiten (2026)
Zitatform
Muehlemann, Samuel (2026): KI in Betrieben. Mehr Ausbildung – aber Weiterbildung zunehmend für anspruchsvollere Tätigkeiten. In: Ifo-Schnelldienst, Jg. 79, H. 03, S. 09-13.
Abstract
"Auf Basis des BIBB-Betriebspanels werden die Folgen der Einführung von Künstlicher Intelligenz für die betriebliche Aus- und Weiterbildung in Deutschland analysiert. Die Ergebnisse deuten darauf hin, dass KI-einführende Ausbildungsbetriebe im Durchschnitt rund 14% mehr neue Auszubildende einstellen. Das spricht für verstärkte Investitionen in den internen Kompetenzaufbau im Zuge des technologischen Wandels. Zugleich verschiebt sich die betriebliche Weiterbildung zugunsten hochqualifizierter Tätigkeiten, während Beschäftigte in Fachkraft- und einfachen Tätigkeiten seltener teilnehmen. Daraus ergibt sich das Risiko einer kumulativen Benachteiligung Geringqualifizierter. Es werden drei wirtschaftspolitische Handlungsfelder abgeleitet: die bessere Integration Jugendlicher mit geringen schulischen Qualifikationen in die duale Ausbildung, eine zielgerichtete Weiterbildungsförderung für Geringqualifizierte sowie schnellere Aktualisierungszyklen für Ausbildungsordnungen und Rahmenlehrpläne." (Autorenreferat, IAB-Doku)
-
Literaturhinweis
Machine learning for labor market matching (2026)
Zitatform
Mühlbauer, Sabrina & Enzo Weber (2026): Machine learning for labor market matching. In: Machine learning with applications, Jg. 23, 2026-02-03. DOI:10.1016/j.mlwa.2026.100861
Abstract
"This paper develops a large-scale machine learning framework to improve labor market matching using rich administrative data. Matching is defined as a job seeker entering employment in a specific occupational field. We exploit comprehensive employment biographies from Germany, covering individual characteristics and job-related information, to estimate employment probabilities across occupations and generate personalized job recommendations. The contribution lies in demonstrating why machine learning methods are particularly well suited for administrative labor market data and outperform traditional statistical approaches. We compare logit, ordinary least squares (OLS), k-nearest neighbors, and random forest (RF). RF consistently achieves the highest predictive performance. Its advantage is rooted in key methodological properties: RF builds an ensemble of decision trees trained on bootstrap samples, introduces random feature selection at each split, and aggregates predictions through majority voting. This enables RF to capture nonlinear relationships and complex interactions, remain robust in high-dimensional settings, and reduce overfitting — features that are particularly relevant for heterogeneous and imbalanced administrative data. Compared to conventional models, RF better exploits the full informational content of employment histories, especially when estimating on all employment spells rather than restricting the sample to unemployment-to-employment transitions. The sample comprises approximately 55 million spells, representing about 6 percent of the German workforce from 2012 to 2018. Our results suggest that ML-based matching, relative to standard statistical approaches, could hypothetically reduce the unemployment rate by up to 0.3 percentage points, highlighting the practical relevance of RF-based decision support for labor market policy." (Author's abstract, IAB-Doku, © Elsevier) ((en))
-
Literaturhinweis
How local labour market skill relatedness and size moderate the impacts of automation (2026)
Zitatform
Njekwa Ryberg, Peter (2026): How local labour market skill relatedness and size moderate the impacts of automation. In: Regional Studies, Jg. 60, H. 1. DOI:10.1080/00343404.2025.2598031
Abstract
"This paper examines how local labour market skill relatedness and size moderate the impacts of automation on occupations across Swedish local labour markets. Using administrative data and a spatially explicit risk of automation measure that accounts for regional differences in occupational task contents, it finds a negative association between automation and employment growth and wage income growth for non-metropolitan occupations between 2011 and 2021. Skill relatedness and labour market size mitigate these negative relationships. In contrast, no negative associations are found for metropolitan occupations. Due to their higher shares of non-automatable tasks, they are more resilient to adverse automation effects." (Author's abstract, IAB-Doku) ((en))
-
Literaturhinweis
Drei Jahre ChatGPT: Auswirkungen von LLMs auf den deutschen Arbeitsmarkt: Teil des Zeitgesprächs "Arbeitszeit im Wandel - Wie sich Wohlstand trotz sinkenden Arbeitskräfteangebots sichern lässt" (2026)
Normann, Morten Grinna;Zitatform
Normann, Morten Grinna (2026): Drei Jahre ChatGPT: Auswirkungen von LLMs auf den deutschen Arbeitsmarkt. Teil des Zeitgesprächs "Arbeitszeit im Wandel - Wie sich Wohlstand trotz sinkenden Arbeitskräfteangebots sichern lässt". In: Wirtschaftsdienst, Jg. 106, H. 4, S. 301-302. DOI:10.2478/wd-2026-0073
Abstract
"Technologischer Wandel ist ein kontinuierlicher Prozess, der Arbeitsmärkte seit jeher beeinflusst. Schon Ökonomen wie David Ricardo warnten offen vor einer Zukunft, in der rasante technologische Entwicklungen (menschliche) Arbeit überflüssig und wertlos machen würden (Ricardo, 1821 in Hollander, 2019). Die neueste Technologie, die als Bedrohung für den Arbeitsmarkt gilt, ist die Entwicklung von Werkzeugen der künstlichen Intelligenz, insbesondere von Large Language Models (LLMs)." (Autorenreferat, IAB-Doku)
-
Literaturhinweis
Human-centred digital transitions and skill mismatches in European workplaces (2026)
Zitatform
Pouliakas, Konstantinos & Giulia Santangelo (2026): Human-centred digital transitions and skill mismatches in European workplaces. (CEDEFOP working paper series / European Centre for the Development of Vocational Training 2026,01), Luxembourg, 163 S. DOI:10.2801/9894877
Abstract
"New digital and artificial intelligence technologies are fast reshaping skill requirements in the EU labour market, fostering skill mismatches. There are marked concerns about the potentially adverse consequences of automation and AI on employment, as well as the lagging competitiveness of EU economies as individuals’ upskilling or reskilling is failing to adapt. To deepen understanding of how digitalisation is affecting the nature of work and skill mismatches in EU labour markets, Cedefop carried out the second wave of the European skills and jobs survey in 2021. In this special edition of Cedefop’s working paper series, ten original, short contributions have been drafted in which researchers explore in depth, for the first time, the ESJS2 microdata. The publication presents a wealth of focused and robust empirical analyses, covering a wide range of different issues on how the digital transition is affecting jobs, skills and training in Europe." (Author's abstract, IAB-Doku) ((en))
-
Literaturhinweis
Mehr Arbeit, weniger Jobs? Konsequenzen der KI-Technologieeinführung: Teil des Zeitgesprächs "Arbeitszeit im Wandel - Wie sich Wohlstand trotz sinkenden Arbeitskräfteangebots sichern lässt" (2026)
Zitatform
Pusch, Toralf, Muhamed Kudic & Akua Franziska Agyepong (2026): Mehr Arbeit, weniger Jobs? Konsequenzen der KI-Technologieeinführung. Teil des Zeitgesprächs "Arbeitszeit im Wandel - Wie sich Wohlstand trotz sinkenden Arbeitskräfteangebots sichern lässt". In: Wirtschaftsdienst, Jg. 106, H. 4, S. 296-300. DOI:10.2478/wd-2026-0072
Abstract
"Steigt durch die zunehmende Einführung von Künstlicher Intelligenz (KI) die Arbeitsbelastung? Und sind Arbeitsplätze bedroht? Daten der WSI-Betriebsrätebefragung zeigen, dass sich KI in ihrer Wirkung – zumindest in der frühen Einführungsphase – von anderen digitalen Technologien zu unterscheiden scheint. Statt einer bei anderen Technologien häufiger auftretenden Arbeitsverdichtung berichteten viele Betriebsräte eher von Arbeitsentlastung, und nach einem Beobachtungszeitraum von zwei Jahren zeigte sich per Saldo ein leichter Stellenaufbau." (Autorenreferat, IAB-Doku)
Aspekt auswählen:
- Gesamtbetrachtungen/Positionen
- Arbeitsformen, Arbeitszeit und Gesundheit
- Qualifikationsanforderungen und Berufe
- Arbeitsplatz- und Beschäftigungseffekte
- Wirtschaftsbereiche
- Arbeits- und sozialrechtliche Aspekte / digitale soziale Sicherung
- Deutschland
- Andere Länder/ internationaler Vergleich
- Besondere Personengruppen