Factors influencing artificial intelligence adoption in human resource management: a meta-synthesis and systematic review of multidimensional considerations Open Access

In today’s digital era, artificial intelligence (AI) is revolutionizing business processes, with human resource management (HRM) experiencing significant transformation through AI’s innovative solutions (Bhardwaj et al., 2020). AI’s evolution, from its industrial origins to contemporary applications, has progressed through the third industrial revolution’s computational advances to today’s sophisticated machine learning applications, fundamentally reshaping organizational practices (Yawalkar, 2019). Defined as an “ideal intelligent machine” capable of environmental adaptation and goal optimization (Sanyaolu and Atsaboghena, 2022), AI has emerged as a cornerstone of Industry 4.0 alongside IoT, big data and cloud technology (Jatobá et al., 2023; Barboza, 2019).

In HRM, AI applications span recruitment, performance management and talent development through machine learning algorithms, natural language processing, and predictive analytics (Rasheed et al., 2024). With 38% of organizations currently utilizing AI and 62% planning adoption (Bhardwaj et al., 2020), the technology has streamlined administrative processes and enhanced operational efficiency (Hossin et al., 2021). However, implementation challenges persist, particularly in developing countries, including financial constraints, employee resistance, job displacement concerns and training requirements (Arslan et al., 2022; Hossin et al., 2021) and data security issues (Jain, 2018).

The human–AI interaction presents additional complexities across various sectors (Arslan et al., 2022), encompassing technology anxiety, trust issues and performance evaluation challenges in hybrid teams (Kaur and Gandolfi, 2023). Successful implementation requires addressing data protection, algorithmic bias (Sabil et al., 2023) and regional-specific challenges such as infrastructure limitations and skill gaps (Chilunjika et al., 2022). While AI offers significant benefits in information processing and decision-making (Alsaif and Sabih Aksoy, 2023; Hossin et al., 2021), understanding adoption factors remains crucial. This study therefore addresses two key questions: (1) What factors influence AI adoption in HRM? (2) How are these factors prioritized in terms of their impact on AI adoption?

HRM has evolved from its industrial revolution origins into a critical driver of sustainable competitive advantage, with pioneering theorists establishing core concepts of selective recruitment and employee training, while the resource-based view emphasizes human capital’s strategic value (Kadam et al., 2022; Anwar and Abdullah, 2021). In response to technological advancement and globalization, modern HRM has embraced digital tools and data analytics to enhance employee engagement and continuous learning (del Val Núñez et al., 2024), while strategic alignment with organizational goals strengthens resilience in volatile environments (Groenewald et al., 2024). The integration of AI has further transformed practices through personalized experiences and predictive analytics, enabling innovation and productivity in multicultural environments while necessitating careful consideration of privacy and inclusivity (Holland et al., 2022; Groenewald et al., 2024; Anwar and Abdullah, 2021).

AI, defined as the simulation of human intelligence, represents a transformative interdisciplinary technology emerging from computer science, cybernetics and mathematical logic to solve complex problems traditionally requiring human cognition (Xu et al., 2021; Liu et al., 2018; Bankins et al., 2024). The field’s evolution since the 1950s, marked by Turing’s foundational question and McCarthy’s term coinage at the 1956 Dartmouth Conference (Zhang and Lu, 2021; Benbya et al., 2020), progressed through stages of cognitive perception, intelligence and decision-making capabilities (Xu et al., 2021). Key milestones include the 1943 artificial neuron model, 1980s neural networks and expert systems, and the 2006 breakthrough in deep neural networks (Zhang and Lu, 2021). In contemporary organizations, AI has evolved from passive tools to active agents, functioning as a disruptive technology that enhances processes through machine learning and data analysis, transforming workforce dynamics and organizational decision-making (Benbya et al., 2020; Xu et al., 2021; Makarius et al., 2020; Robert et al., 2020).

AI’s integration into HRM has revolutionized organizational practices by merging multidisciplinary elements of mathematics, management, computer science and linguistics to enable complex data analysis and pattern recognition in key processes, from applicant ranking to hiring efficiency (Kshetri, 2021; Vasantham, 2021; Qiu and Zhao, 2018). This technology empowers human resource (HR) professionals to focus on strategic initiatives while leveraging AI-driven analytics for decision-making and employee assessment, ultimately creating dynamic work environments guided by real-time data and personalized experiences (Singh et al., 2023; Kaushal et al., 2023; Chukwuka and Dibie, 2024). However, successful implementation necessitates addressing workforce displacement, upskilling needs and ethical considerations to maintain equilibrium between technological innovation and human-centric approaches (Singh et al., 2023).

AI has transformed HRM through the automation of routine tasks, including resume scanning and candidate evaluation, enhancing efficiency while minimizing human bias in decision-making (Vasantham, 2021; Qiu and Zhao, 2018; Vrontis et al., 2023). However, despite these transformative benefits, organizations face substantial implementation challenges. Most critically, job displacement emerges as a primary concern, with projections indicating automation could eliminate up to 50% of existing positions in the coming decades (Abdeldayem and Aldulaimi, 2020). Furthermore, employee resistance and technology anxiety create additional obstacles, necessitating organizations to develop collaborative environments where AI serves as an enabling tool rather than a threat (Suseno et al., 2021). In this context, the integration of AI systems and humanoid service robots requires strategic workforce role transitions (Vrontis et al., 2023). Consequently, successful implementation hinges on achieving equilibrium between technological advancement and workforce preparedness through targeted initiatives addressing skill gaps, cultural transformation and organizational alignment (Suseno et al., 2021; Vasantham, 2021).

This study employs a distinctive meta-synthesis approach to analyze AI adoption factors in HRM, systematically integrating both qualitative and quantitative findings to generate comprehensive insights (Nye et al., 2017). The methodology’s strength lies in its interpretive and dialectical approach, enabling “third-order” interpretations through three key mechanisms: first, reciprocal translation across diverse studies; second, thematic synthesis of emerging patterns and third, analytical refinement of theoretical constructs (Nye et al., 2016). This rigorous process reveals the underlying mechanisms while emphasizing analytical transferability over mere generalizability, effectively bridging the paradigmatic divide between research and practice. The research systematically implements Sandelowski and Barroso’s (2007) comprehensive six-step framework, progressing from systematic search and evaluation through synthesis and interpretation, ensuring methodological rigor while maintaining interpretive depth (Lachal et al., 2015).

The meta-synthesis implementation follows distinct stages as shown in Figure 1. The research question formulation, as the first step, considers key parameters including methodology, study population, time frame and subject matter. The foundational questions outlined in Table 1 guide the research direction and establish the study’s framework.

In the second step, a comprehensive systematic review was conducted using keywords related to AI and HRM. The search covered the Science, Google Scholar, Scopus, Emerald and Springer databases, selected for their extensive coverage, credibility, search capabilities and access to current research in management, technology and human sciences. Keywords included “Artificial Intelligence,” “Human Resources Management,” “Adoption” and “Acceptance.”

After completing the search for articles intended for meta-synthesis, as seen in Table 2, the following process was undertaken to select articles. Initially, all articles related to the topic, totaling 87, were downloaded from databases. Subsequently, a thorough and purposeful five-stage refinement process was conducted.

From the initial 87 articles extracted from reputable databases.

1. Elimination of duplicates: In the first step, 15 duplicate articles (17%) were removed to ensure originality and uniqueness of the selected studies.

2. Title screening: The titles of the remaining articles were carefully reviewed, leading to the exclusion of 22 articles (25%) that were deemed irrelevant to the topic of AI adoption in HRM.

3. Abstract review: Abstracts of the remaining articles were examined to assess relevance and methodological rigor. This stage resulted in the removal of 19 additional articles (21%) that failed to meet the inclusion criteria.

4. Full-text content analysis: A deeper evaluation of the full texts was conducted to identify studies with insufficient information or lacking direct relevance to the research topic. Consequently, 11 more articles (12%) were excluded.

5. Quality assessment using Critical Appraisal Skills Programme (CASP) criteria: Quality assessment of the final 20 articles was conducted using the CASP checklist (Lachal et al., 2015). The evaluation covered 10 key criteria: (1) clarity of research aims, (2) appropriateness of methodology, (3) relevance of research design, (4) recruitment strategy suitability, (5) data collection adequacy, (6) researcher–participant relationships, (7) ethical considerations, (8) data analysis rigor, (9) findings clarity and (10) research value. Each criterion was scored from 0 to 5 points. All studies met quality standards, with most scoring highly across criteria, as detailed in Table 3.

In the fourth stage, the analysis focused on identifying and categorizing factors influencing AI adoption in HR processes across organizational, technological and human contexts. This systematic process enabled the extraction of key concepts and patterns, forming the basis for developing a conceptual framework.

The fifth stage revealed that no previous study had taken a comprehensive approach to AI adoption in HR processes. A systematic literature review demonstrated that earlier research had primarily focused on specific aspects of this topic. Table 4 displays the final extracted codes for each dimension, category and influencing factor, providing deeper insights into the studied phenomenon and establishing groundwork for an integrated model of AI adoption factors in HR.

To validate research reliability, Cohen’s Kappa coefficient (1960) was employed with two independent evaluators classifying articles into five main categories of AI adoption factors in HR (organizational-strategic, technological-operational, human-centric, challenges-considerations and environmental-economic). The Kappa coefficient of 0.81 demonstrated strong inter-rater agreement, confirming high validity in both categorization and factor extraction processes.

The final stage integrates the research outcomes into a cohesive process, systematically synthesizing findings from previous studies on AI adoption in HR processes. Through detailed analysis and classification, this culminated in a comprehensive model (Figure 2) that illustrates the factors influencing AI adoption in HR, providing both a research summary and a conceptual framework.

Figure 3 ranks the factors influencing the adoption of AI in HRM based on theme frequency. These data not only demonstrate the relative importance of each factor but also provide a comprehensive overview of the influential factors in this domain.

Through a meta-synthesis of 2015–2023 studies, this research examined AI adoption factors in HRM. The analysis identified five main themes: organizational and strategic factors, technological and operational factors, human-centric factors, challenges and opportunities and environmental and economic factors.

Analysis highlights ethical and legal considerations as the primary concern, with the highest frequency (16), emphasizing the critical importance of addressing risks and regulatory challenges in AI implementation. Notably, economic factors demonstrated the lowest frequency (3), suggesting that noneconomic considerations currently dominate organizational decision-making.

Organizational and strategic elements emerge as foundational, with leadership factors underlining the importance of vision. Psychological and attitudinal factors demonstrate that successful adoption depends heavily on employee perceptions and organizational culture. Operational benefits and technical infrastructure provide the technological groundwork for implementation, while management challenges highlight integration complexities.

This analysis reveals that AI adoption in HRM requires a holistic approach balancing technological capabilities with human readiness, ethical considerations and strategic opportunities. Success depends on creating an organizational ecosystem that effectively leverages technology while maintaining human-centric values and compliance standards.

The findings demonstrate that multiple interconnected factors shape AI adoption in HRM, with ethical and legal considerations emerging as the predominant factors (16 occurrences). Singh and Pandey (2024) recount a case in which a global company discontinued AI-based hiring in 2018 due to biases with significant legal and societal impacts. Parallel to these legal considerations, ethical frameworks play a crucial but separate role, as emphasized by Asif’s (2024) research on developing ethical guidelines for AI implementation in recruitment processes.

The multifaceted nature of AI adoption is particularly evident in the domain of user trust, as identified by Kelly et al. (2023) in their comprehensive analysis. While ethical and legal compliance forms one dimension of trust-building, the phenomenon is inherently more complex, encompassing multiple interconnected factors: technological reliability, algorithmic transparency, system performance metrics, user experience outcomes and demonstrated operational efficiency (Xu et al., 2024; Chatterjee et al., 2024). This multidimensional trust framework is further complicated by privacy considerations, which Choi (2021) identifies as a significant moderator of user adoption willingness. Tuffaha and Perello-Marin’s (2023) identification of research gaps in ethical implications suggests an emerging scholarly focus on these critical aspects.

Organizational and leadership factors, alongside psychological and attitudinal considerations (13 occurrences each), represent the second tier of influence in AI adoption dynamics. Alsheibani et al.’s (2020) empirical investigation demonstrates the crucial interplay between top management support, organizational readiness and regulatory compliance. This finding is reinforced by Pillai and Sivathanu’s (2020) analysis of leadership support’s role in successful AI integration within talent acquisition processes. Hamm and Klesel’s (2021) identification of 13 distinct organizational factors further validates the complexity of institutional dynamics in AI adoption. The psychological dimension, particularly evident in Ochmann and Laumer’s (2020) examination of candidate expectations in automated recruitment, reveals the intricate relationship between user attitudes and adoption outcomes, a finding substantiated by Kelly et al.’s (2023) predictive modeling of AI acceptance patterns.

Operational benefits and implementation challenges (11 occurrences each) constitute the next significant cluster of factors. Cost-effectiveness and comparative advantages emerge as primary drivers of adoption (Pillai and Sivathanu, 2020), while performance expectancy significantly influences implementation success (Kelly et al., 2023). However, organizations face substantial challenges, including resource allocation demands, organizational resistance, standardization complexities (Rane et al., 2024) and cybersecurity considerations (Pillai and Sivathanu, 2020).

Technical and infrastructure factors (ten occurrences) play a crucial role in adoption outcomes, with vendor support emerging as a critical success factor (Pillai and Sivathanu, 2020). The environmental and economic context (nine occurrences) shapes adoption through regulatory frameworks (Alsheibani et al., 2020) and market competition dynamics (Pillai and Sivathanu, 2020; Rane et al., 2024).

The findings largely align with existing literature while highlighting emerging trends, particularly the prominence of ethical and legal considerations in AI adoption for HRM. The high ranking of both organizational/leadership and psychological/attitudinal factors emphasizes that successful AI integration requires a holistic approach beyond technical implementation. This meta-synthesis demonstrates the multidimensional nature of AI adoption in HRM, suggesting organizations must address these interrelated factors comprehensively to optimize AI implementation while managing associated challenges.

This study’s findings align with and extend key technology adoption frameworks. The prominence of psychological and attitudinal factors corresponds with the Technology Acceptance Model’s core constructs (Davis et al., 1989), while the high frequency of ethical and legal considerations (16 occurrences) suggests the need to expand traditional frameworks. This aligns with recent developments in TAM2 ^[1] and UTAUT ^[2] (Mogaji et al., 2024), which recognize broader social and psychological determinants.

The findings also reflect the five characteristics of the diffusion of innovation theory (Muliadi and Usman, 2024). Organizational readiness and strategic alignment mirror the theory’s emphasis on compatibility, while ethical considerations highlight innovation complexity. This research uniquely bridges the technology acceptance model’s cognitive-psychological focus with the diffusion of innovation theory’s organizational–societal perspective, demonstrating how individual user experiences interact with broader organizational dynamics in AI adoption for HRM.

The prominent role of ethical and legal considerations introduces a critical dimension often understated in traditional frameworks. This suggests the need for an integrated adoption framework that synthesizes insights from existing theories while incorporating ethical responsibility, human-centric readiness and strategic alignment. Such theoretical advancement would better address the complexities of AI adoption in HRM and guide future research on the interplay between technological, psychological and organizational factors.

Successful AI adoption in HRM requires a comprehensive strategy addressing multiple dimensions, with ethical and legal considerations emerging as the highest priority factor. Organizations must establish robust frameworks for data privacy, algorithmic fairness and regulatory compliance to mitigate the risks of bias and discrimination.

Organizational leadership and psychological factors form the next critical layer, requiring strong strategic alignment and cultural adaptation. This involves implementing transparent communication strategies, developing targeted training programs and actively engaging employees to address technology resistance and build trust. These initiatives should leverage AI’s operational benefits while managing implementation challenges through systematic change management approaches.

Technical infrastructure and strategic advantages demand equal attention, requiring scalable systems and expert collaboration to optimize AI performance. Environmental and economic factors must be considered within the broader ecosystem, while organizational strategies should focus on enhancing employee experience and improving decision-making processes.

The successful integration of AI in HRM ultimately depends on balancing these interconnected dimensions while prioritizing factors based on their empirically determined significance. This evidence-based approach ensures organizations can effectively leverage AI’s transformative potential while maintaining ethical responsibility and operational excellence.

Future research should address the critical issue of AI bias, which remains a significant challenge to fairness and transparency in HRM. Bias in AI systems often stems from imbalances or prejudices in training data, algorithmic limitations or design flaws, potentially leading to discriminatory practices and undermining trust in AI-based decisions (Ahn et al., 2022; Leong and Sung, 2024). Exploring strategies for detecting, mitigating, and preventing such biases is crucial for ensuring ethical and equitable AI adoption in HR processes. Addressing these concerns will not only improve the ethical implementation of AI but also enhance its acceptance and effectiveness in diverse organizational contexts.