The Meta-HR framework: a systematic literature review on AI-accelerated HR systems for digital talent transformation

Abstract

With the proliferation of AI-based tools in human resource management (HRM), the literature remains fragmented, addressing recruitment automation, performance analytics, and algorithmic management separately. There is a lack of a solid theoretical framework integrating organizational, technological, and human aspects for digital talent transformation. This paper proposes the Meta-HR Framework, a closed, self-optimizing architecture that reconceptualizes AI as a strategic driver of socio-technical system transformation. Based on PRISMA, a systematic literature review of 119 peer-reviewed articles (2020–2025) from Scopus, Emerald, and ScienceDirect was conducted. The Extended TOE framework, multilevel systems theory, and dynamic capabilities guided the thematic synthesis. Four interdependent layers comprise the Meta-HR Framework: (1) AI Infrastructure, which consolidates data, Machine Learning models, and Natural Language Processing engines; (2) Cognitive HR, which translates insights into AI-enhanced decision-making systems; (3) Transformational HR, which embeds digital fluency, capability development, and workforce agility; and (4) Meta-Learning Loop, which closes the system through reinforcement learning and capability development. Five propositions are derived. This four-layer risk assessment checklist helps HR leaders develop ethical, human-centered, and adaptive AI-integrated HR systems while managing algorithmic bias, skills gaps, and resistance to change. The Meta-HR Framework offers a closed, dynamic architecture based on socio-technical systems theory; it synthesizes TOE, multi-level systems theory, and dynamic capabilities into a self-evolving HR system. This contrasts with AI-HRM maturity models that depict gradual adoption. According to the SLR analysis, no previous model has integrated these three theoretical perspectives into a closed architecture.