New Preprint: Perceived Industry Resilience to AI

A new preprint with Birte Keller and Frank Marcinkowski is now available on SocArXiv: “Perceived Industry Resilience to AI: A Two-Wave Survey of Employees Across 15 Sectors.”

Artificial intelligence is expected to transform industries unevenly, yet little is known about how employees themselves evaluate the capacity of their own industry to adapt. The paper introduces perceived industry resilience to AI as a meso-level construct linking individual perceptions to sectoral conditions, and examines its level, sectoral differences, short-term change, and individual-level correlates.

The study draws on a two-wave repeated cross-sectional survey of employed respondents in Germany (T1: July 2025; T2: January/February 2026; pooled N = 1,707) covering 15 industry sectors, with research questions and hypotheses preregistered between the two waves.

Perceived resilience to AI was generally low and showed no evidence of overall change between waves. Sectors differed markedly: employees in information technology and data processing reported the highest resilience perceptions, while those in administration, public service, and social services reported the lowest. Employees who regarded the benefits of AI for their own profession as outweighing its risks, and who saw political institutions as responsive, reported higher perceived resilience in both waves. The findings point to occupational and political evaluations as anchors of resilience perceptions, with implications for sector-specific workforce policy and communication.

The preprint, preregistration, and supplemental materials are available via SocArXiv.