AI Workforce Architecture and Digital Innovation: Complementarity, Governance, and the Value of Balance
Type: article
Abstract: Firms with comparable artificial intelligence headcounts can distribute that talent very differently across the functions AI work requires. We distinguish Invention, Application/Design, and Scaling/Governance roles and link Revelio Labs workforce data to Compustat and KPSS patent data for 3,236 publicly traded U.S. firms from 2012 to 2022. In firm and year fixed-effects Poisson pseudo-maximum likelihood (PPML) models, the Invention × Application/Design staffing interaction is associated with broader digital technology scope and entry into new digital subclasses, but not with greater patent volume. The relationship with forward citations is stronger where Scaling/Governance staffing is higher, and firms with more balanced role composition exhibit broader exploratory search. Alternative taxonomies and specifications yield the same pattern. For research on digital innovation and AI capability, this locates one source of innovation heterogeneity in the architecture of the AI workforce: which functions a firm covers, and how they combine.
Authors: Schumacher, Christian, Tihanyi, Can
Institution: Vienna University of Economics and Business
Language: English
Subfield: Strategy and Management
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7222991172