Working Paper No. 370: AI Unboxed and Jobs: A Novel Measure and Firm-Level Evidence from Three Countries - Ratio

Working Paper No. 370: AI Unboxed and Jobs: A Novel Measure and Firm-Level Evidence from Three Countries - Ratio

Liknande innehåll

Who Adopts AI? Evidence on Firms, Technologies and Workers

Pulito, G., Pytlikova, M., Schroeder, S. & Lodefalk, M.

Publiceringsår

2026

Publicerat i

GLO Discussion Paper

Sammanfattning

Using two waves of nationally representative Danish firm surveys linked to employer–employee administrative registers, we study how adoption varies across artificial intelligence (AI) and related advanced technologies. We show that AI adoption is highly technology-specific. While firm size and digital infrastructure predict adoption broadly, workforce composition operates through distinct channels: STEM-educated workforces predict core AI adoption, whereas non-STEM university-educated workforces are associated with generative AI adoption, indicating different human capital complementarities. The factors associated with adoption differ from those predicting deployment breadth: firm size and digital maturity matter for both, whereas workforce composition primarily predicts adoption alone. Machine learning and natural language processing are deployed across multiple business functions, whereas other advanced technologies remain concentrated in specific operational domains. Individual-level evidence provides a foundation for these patterns, with awareness of workplace AI usage concentrated among managers and high-skilled workers. Self-reported AI knowledge is higher among younger and more educated individuals. Finally, commonly used occupational AI exposure measures vary substantially in their ability to predict observed adoption, with benchmark-based measures outperforming patent-based and large-language-model-focused alternatives. These findings show that treating AI as a monolithic category obscures economically meaningful variation in who adopts, what they deploy, and how well existing measures capture it.

Who is afraid of AI? Who should be?

Engberg, E., Görg, H., Hellsten, M., Javed, F., Lodefalk, M., Längkvist, M., & ..

Publiceringsår

2026

Publicerat i

Kiel Policy Brief, 2026.

Sammanfattning

Occupations that are highly cognitive, non-physical, and low in social interaction — typically higher-skill white-collar roles such as data analysts, software developers, and translators — turn out to be highly AI-exposed

Occupations requiring manual dexterity or intensive interpersonal contact — such as construction labourers or nursing aides — remain among the least exposed to current AI technologies

Aggregate occupational exposure to AI has risen markedly since 2010, with especially rapid gains in the late 2010s and early 2020s

Our baseline estimates show no detectable effect of AI exposure on total firm employment, while it is associated with clear skill upgrading

Engberg, E., Görg, H., Hellsten, M., Javed, F., Lodefalk, M., Längkvist, M., & .. (2026). Who is afraid of AI? Who should be?. Kiel Policy Brief, 2026.

The Impact of AI on the Labour Market: Essays on Transformative Technology, Occupations, and Firms

Publiceringsår

2026

Publicerat i

Örebro University.

Sammanfattning

The topic of this thesis is the economics of transformative technology, with the impact of artificial intelligence (AI) on the labour market as the primary focus.

Analysing German data, Essay I shows that occupational AI exposure was associated with wage gains, and an increased focus on knowledge-intensive tasks. There is a clear contrast between the types of work that are exposed to AI, versus robotics.

Essay II finds that AI exposure is associated with AI adoption and increased labour demand, as measured by job vacancy postings, in Swedish establishments/workplaces.

Essay III develops a novel measure of occupational AI exposure, called Dynamic AI Occupational Exposure (DAIOE). AI exposure is shown to be associated with upskilling at the firm level in Sweden, Denmark, and Portugal.

Essay IV analyses the labour market implications of the growing social and verbal capabilities of large language models (LLMs). Analysis of occupational data from O*NET and job ads provides a map of the most important types of social work tasks. Among social tasks, verbal communication tasks have the strongest association with occupational exposure to LLMs.

Essay V is about the impact of venture capital (VC) on start-up firms. Investment from both private and governmental VCs is found to increase sales with a 2-3 year delay, driven primarily by efficiency gains, and to some extent, capital investment. Governmental VCs are more likely to make follow-on investments in non-growing firms.