Artificial Intelligence in the Labor Market: Evidence on Worker Inclusion, Exclusion, and Discrimination—A Systematic Review

Artificial Intelligence (AI) is increasingly used in recruitment, performance management, and algorithmic work management, with potentially divergent implications for worker inclusion, exclusion, and discrimination. This systematic review synthesizes peer-reviewed evidence on (i) which AI applications in labor-market settings are linked to inclusion/exclusion outcomes, (ii) the mechanisms and contextual moderators shaping these effects, and (iii) governance and human-resource management responses proposed in the literature. Guided by PRISMA 2020, we searched Scopus and Web of Science (Title/Abstract/Keywords) for English-language journal articles published between 2015 and 2025. Nineteen studies met the eligibility criteria and were analyzed using qualitative thematic synthesis. The evidence indicates an ambivalent pattern: AI can support inclusion through assistive technologies and improved matching, but it can also exacerbate occupational polarization, digital exclusion, and discriminatory outcomes when models are trained on biased data or deployed without transparency and accountability. Outcomes depend on complementary organizational practices, workers’ access to skills, and the regulatory environment. Based on an evidence map of the included studies, we propose a hybrid governance model combining technical and organizational audits, inclusive upskilling/reskilling, participatory regulation, and responsible HR policies to align AI innovation with decent and inclusive work. Given the focused Title/Abstract/Keywords query and the small, heterogeneous corpus, the findings are interpreted as a scoped evidence map rather than an exhaustive census of all AI-and-work research. The model’s contribution lies in integrating four interdependent governance layers—technical, organizational, workforce, and regulatory—within a single labor-market framework. Accordingly, the review should be read as a focused qualitative evidence synthesis, and the proposed model as an evidence-informed conceptual framework that warrants future empirical validation.