AUSPICE · AI Usage Strategies: Policy, Institutions and Consequences in Europe

Lines connect the coordinator with its partners.

Project objective

AUSPICE investigates why some countries are more inclined to implement artificial intelligence (AI) to complement workers, while others tend to automate tasks. It also examines how these choices influence jobs, inequality, and public support for AI. Moving beyond simple AI adoption rates, it measures “AI usage strategies” and their impacts, focusing on a key European priority: aligning technological progress with social inclusion through education and lifelong learning. The project has three objectives: (1) develop a longitudinal country‑ and sector‑level index that differentiates automation from augmentation; (2) identify the political and institutional factors driving these strategies; and (3) evaluate their distributional and political consequences. The methodology combines latent‑variable modelling to develop the index using harmonised EU sources (e.g., Eurostat, Eurofound), panel econometrics to explain differences across countries and sectors, as well as the relationships between strategies and group-level outcomes (e.g. gender, education, and occupation), and a pre‑registered online experiment to see how augmentative versus automating frames influence support for AI and related policies. Expected outcomes include a validated public index of AI usage strategies, evidence on policy solutions—such as skills formation systems, coordination bodies, and social partnerships—that promote augmentation, and guidance for policies that boost productivity while limiting inequality. The research outputs will include three working papers, a policy brief, and accessible communications via UNIBO platforms and the EUFutures Jean Monnet Centre of Excellence. All preprints will be uploaded to OSF, the new index and reproducible codes will be shared under open licences on Zenodo, and all workflows will be publicly available on GitHub. Training will feature lectures on latent variable models and Bayesian methods, and knowledge exchange with an interdisciplinary AI scholarship.

Original text from CORDIS.

Participants

• ALMA MATER STUDIORUM - UNIVERSITA DI BOLOGNA · BolognaCoordinatorItaly

Links

• View on CORDIS
• DOI: 10.3030/101280955