Artificial Intelligence and the Teaching Workforce: Evidence from National Survey Data in the United States

• In Event: Sociology of Education Section Roundtables
  In Refereed Roundtable (60 minute): Table 16. AI for Teaching and Learning

Abstract

Artificial intelligence (AI) is moving from novelty to governance in U.S. schools, yet we lack nationally representative evidence on how AI intersects with teachers’ work, capacity, and well‑being. Using the TALIS 2024 survey, the most recent Teaching and Learning International Study released in late 2025, this study examines how AI‑specific professional development (PD), belief orientations, usage intensity, and perceived barriers relate to instructional fulfillment and occupational well‑being. Guided by policy implementation, professional capital, and job demands‑resources perspectives, this study estimates design‑based linear models that implement balanced repeated replication with final teacher weights and report survey‑weighted statistics.
Findings show two in five U.S. teachers reported AI‑focused PD and AI use. Positive beliefs about AI are associated with higher job satisfaction and better workplace well‑being, though also with slightly higher workload stress; critical beliefs are associated with higher self‑efficacy and fulfillment but lower satisfaction/joy and poorer well‑being. Perceived barriers (e.g., insufficient support, unclear guidance, feeling overwhelmed) predict lower satisfaction and joy and higher stress, while greater usage breadth correlates with poorer well‑being and higher stress net of covariates. In contrast, general (non‑AI) PD breadth and professional collaboration show consistent protective associations (higher fulfillment/efficacy/satisfaction; lower stress). Weekly hours are strongly linked to strain.
This study argues AI policy is workforce policy: without time‑saving deployments, clear guardrails, and collaborative infrastructure, expanding AI across tasks risks technostress, even when teachers maintain instructional efficacy. The paper contributes nationally representative evidence to debates on school governance, teacher labor, and technology adoption, offering concrete implications for district and state policy (time audits, procurement that prioritizes net time relief, risk‑aware PD nested within broader capacity building, and scheduled collaboration time). The study aligns with ASA’s 2026 theme on disrupting the status quo by centering conditions under which AI is worth using for teachers, not just whether it is used.

Author

• Lei Jiang, University of Kansas