Digital human resource management in the era of artificial intelligence: research performance and future research agenda

In today's volatile business context, digital transformation is increasingly recognized as an essential strategy for all organizations to maintain competitiveness. No longer limited to the adoption of new technologies, digital transformation and artificial intelligence exert a profound influence on all aspects of management, encompassing strategic planning, resource organization, leadership, operational performance monitoring (Sheveleva et al., 2021) and employee training (Passmore et al., 2025). To meet the challenges and seize the opportunities of the digital age, organizations are compelled to restructure internal processes, rethink management mindsets and enhance their digital capabilities.

Amid the widespread wave of digital transformation, digital human resource management (DHRM) has emerged as a key factor influencing organizational performance and adaptability. The rise of DHRM has been driven not only by health crises like COVID-19 but also by inevitable technological advancements, which are fundamentally reshaping how organizations recruit, train, evaluate and retain talent. Integrating digital technologies into human resource (HR) strategies enables organizations to better align business objectives with human capital capabilities, thereby strengthening their competitive edge in an increasingly dynamic and complex digital economy (Shahiduzzaman, 2025).

Despite the growing importance and rapid adoption of DHRM, there have been few comprehensive literature reviews on the subject. For example, Ghazy and Fedorova (2022) focused only on the hospitality industry with outdated data. Bindra et al. (2025) limited their scope to 2015–2024 using multiple disparate databases. Although many organizations have rapidly implemented DHRM to gain a competitive edge, there remains ambiguity regarding the deeper implications and long-term consequences of this transformation (Bindra et al., 2025). There are still gaps in understanding how AI can support and enhance rather than entirely replace human-led HRM functions (Bindra et al., 2025). Moreover, future research directions closely related to the context of emerging technologies appear not to have been proposed in previous review studies. Existing studies have also yet to show how DHRM has conceptually shifted from efficiency-focused digitalization to strategic transformation through AI and socio-technical integration. Therefore, an up-to-date review of the literature is urgently needed, covering the entire body of research from the earliest studies to the present.

The overarching aim of this paper is to conduct a comprehensive literature review of the field of DHRM. Drawing on data from the Scopus database from the first recorded study in 2004–2025 and employing bibliometric analysis, this research seeks to evaluate a large body of literature and visually and systematically map the intellectual structure of the field (Donthu et al., 2021). Specifically, the study addresses the following research questions:

This study makes several key contributions to the DHRM literature. First, it provides a comprehensive and up-to-date bibliometric analysis of the field, capturing its entire evolutionary path from its inception (2004) to 2025, thereby addressing the limitations of previous reviews that relied on older or narrower datasets. Second, by systematically employing bibliographic coupling and keyword co-occurrence analysis, it identifies and visualizes the intellectual structure and core themes that have shaped the domain, offering a clearer map of its theoretical foundations and research fronts. Beyond updating datasets and visual mappings, this study offers a conceptual synthesis by revealing that DHRM research has undergone a shift, from viewing digitalization as operational efficiency to understanding it as a driver of strategic transformation, ethical tension and socio-technical integration. Third, this study goes beyond performance analysis by identifying timely future research agendas, particularly acknowledging the growing role of AI, alongside excessive digitalization, humanoid robots, ethical considerations and the need for multi-contextual research.

The remaining sections of this study are structured as follows: Theoretical Background, Method, Findings and Discussion, Future Research Agenda, Implications and Conclusion.

The term DHRM describes the use of digital tools and technologies to handle human resource tasks such as recruitment, career development, training, performance management and compensation (Prabhu et al., 2023). It is considered an evolution of earlier technology-based human resource management (HRM) concepts (Strohmeier, 2020).

The transition from Electronic Human Resource Management (e-HRM) to DHRM reflects a shift from operational digitalization to strategic transformation in HRM. Since the 1980s, the development of information technology has enabled HR to evolve from an administrative function to a strategic partner, thereby laying the foundation for e-HRM (Dulebohn and Stone, 2018). E-HRM is typically understood as a web-based HR system that automates processes such as recruitment, training and performance management, thereby enhancing efficiency and real-time information access to information (Prabhu et al., 2023).

However, DHRM has evolved beyond e-HRM by integrating advanced technologies such as AI, machine learning and cloud computing. These technologies not only improve operational efficiency but also support strategic decision-making and enhance employee experience (Prabhu et al., 2023). The key distinction lies in the fact that while e-HRM primarily digitizes HR activities, DHRM emphasizes a shift toward higher-value strategic functions (Prabhu et al., 2023). This transformation is further enabled by cloud-based e-HRM systems and digital HR tools such as chatbots, people analytics and human capital management systems (Sharma and Sengupta, 2024).

Despite significant progress, DHRM research remains fragmented around several interrelated tensions rather than a single dominant theoretical perspective. These tensions concern not only the adoption of digital technologies in HR functions but also the changing strategic role of HR, the socio-technical conditions shaping implementation and the ethical implications of AI-enabled decision-making. Reviewing these debates is important because they reveal how the field has moved from a narrow concern with HR process digitalization to broader questions of strategic transformation, responsible AI governance and sustainable human-technology integration.

A primary discussion concerns whether DHRM merely represents the digitalization of existing HR functions or a profound strategic transformation (Strohmeier, 2020; Bondarouk and Brewster, 2016). Recent studies suggest that DHRM increasingly serves as a source of strategic value through algorithmic management, data analytics and AI, enhancing both efficiency and predictive decision-making capabilities (Angelova and Anguelov, 2025). However, this leads to a further debate regarding the relationship between technology and organizations: while technology can reshape processes and power structures, organizational culture, digital competence and human involvement ultimately determine how technologies are implemented and adopted. This highlights the relevance of a socio-technical perspective rather than a purely technological determinist approach.

At the same time, contemporary debates increasingly focus on the dual roles of AI and algorithmic systems. While they enhance efficiency and objectivity, they also raise concerns regarding bias, privacy and the dehumanization of the workforce (Angelova, 2025). Issues related to transparency, explainability and data quality remain insufficiently addressed (Setiawati et al., 2025), giving rise to the debate over whether DHRM should prioritize short-term performance or long-term sustainability. Several studies caution that an excessive focus on immediate efficiency may undermine employee trust, well-being and organizational culture in the long run. Consequently, current trends emphasize the development of hybrid, human-centered AI governance models that seek to balance technological efficiency with social responsibility (Setiawati et al., 2025).

In addition, emerging debates address the degree of humanization and contextual applicability of DHRM. Heavy reliance on AI may diminish emotional and social interaction elements in HRM, potentially affecting employee engagement and trust (Chao et al., 2025; Yavuz and Ateş, 2026), while trust in automated systems remains limited without adequate human oversight. Furthermore, the generalizability of DHRM models is increasingly questioned, as most studies focus on developed economies; factors such as digital infrastructure, data capabilities and organizational culture may significantly influence implementation outcomes across different contexts. The rapid pace of technological change also renders many frameworks quickly outdated, intensifying debates regarding the long-term stability and adaptability of DHRM (Chao et al., 2025).

These debates indicate that DHRM should be understood as a complex socio-technical transformation rather than a simple extension of e-HRM or HR process automation. The rise of AI has intensified this transformation by expanding the analytical and predictive capabilities of HR systems while simultaneously creating new concerns regarding transparency, fairness, accountability and employee well-being. However, prior studies remain fragmented across separate discussions of technology adoption, strategic value, employee outcomes and ethical risks. This fragmentation limits the current understanding of how DHRM has evolved intellectually and how AI is reshaping its research agenda. Therefore, a comprehensive bibliometric and content-based review is needed to map the field's knowledge structure, integrate its major debates and identify future directions for DHRM research in the AI era.

This study adopts a bibliometric review design, which is considered appropriate for synthesizing large bodies of literature, identifying conceptual structures and mapping the evolution of emerging research fields (Donthu et al., 2021). Bibliometric analysis provides a systematic and replicable approach to assessing scholarly performance and uncovering intellectual patterns based on citation, authorship and keyword relationships (Hallinger and Kovačević, 2019).

Scopus was selected due to its comprehensive coverage of peer-reviewed publications in management and its suitability for bibliometric mapping (Hallinger and Kovačević, 2019). It offers broader indexed coverage in the management field than alternative databases (Santos et al., 2023). Focusing on a high-impact database such as Scopus ensures access to authoritative literature reflecting the most influential studies in the field (Singh et al., 2025). While relying on a single database may limit scope, this trade-off prioritizes data consistency and avoids duplication inherent in merging multiple sources.

The keyword list was developed through a four-step procedure. First, core conceptual terms, including “digital human resource management”, “DHRM”, “e-HRM”, “electronic HRM”, “electronic human resource management” and “virtual human resource management”, were derived from prior literature (e.g. Strohmeier, 2020; Prabhu et al., 2023; Sharma and Sengupta, 2024). Second, we expanded the list by reviewing keywords from recent articles in the field. Technology-related terms were added to capture system-level digitalization. Third, function-specific terms were incorporated to ensure coverage of core HR processes. Fourth, the consolidated list was reviewed by two senior HR technology managers to ensure practical relevance. The final keywords are presented in Table 1.

The data screening process adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Moher et al., 2009) to enhance transparency and replicability. PRISMA emphasizes the implementation of comprehensive literature search strategies, detailed data extraction, risk of bias assessment and structured reporting. This ensures that evidence synthesis is reliable and reproducible, allowing systematic reviews and meta-analyses conducted according to PRISMA to achieve high quality and credibility (Kumar et al., 2023). Figure 1 illustrates the four steps of PRISMA.

First, the keywords were searched within the title, abstract and keywords fields on Scopus, without applying any screening criteria initially. This search yielded 1,400 documents.

Next, inclusion and exclusion criteria were applied to refine the dataset and enhance its relevance, following Donthu et al. (2021), as detailed in Table 1. No time restriction was imposed to capture the full scope of the literature from the field's first publication up to 20 May 2025 and to address gaps in previous reviews. The study was limited to the “business, management and accounting” subject area to maintain focus, given the large initial dataset (1,400 documents) and the presence of many technology-related studies (e.g. computer science). As Donthu et al. (2021) emphasize, narrowing the scope is necessary to reduce complexity and keep the review on-topic. This process left 359 documents. For interdisciplinary topics with broad keyword strategies, a high exclusion rate during screening is common (Hallinger and Kovačević, 2019).

The remaining documents were manually screened by the authors, who divided the workload by reviewing titles, abstracts and, when necessary, full texts, followed by cross-checking their results. This process enhanced the rigor and accuracy of the dataset (Donthu et al., 2021). Irrelevant studies were excluded, resulting in a final sample of 328 documents for analysis.

The study employed both performance analysis and science mapping techniques to provide a holistic understanding of the DHRM field. VOSviewer 1.6.20 was used for visual analyses due to its effectiveness in handling large bibliometric datasets and generating clear network maps (Donthu et al., 2021). Excel was used to produce descriptive statistics of the field.

To analyze the publication growth of the field, the dataset was sorted by publication year, and the COUNTIF function in Excel was applied. To analyze country contributions, VOSviewer was employed using the co-authorship technique, with countries as the unit of analysis and the full counting method applied. The data were exported from VOSviewer in text format, then converted into Excel and sorted in descending order by the number of publications the number of citations and country name in alphabetical order.

The identification of key themes followed three main steps. First, bibliographic coupling was conducted using VOSviewer, with documents as the unit of analysis (Koseoglu et al., 2016). A threshold of 280 documents was selected based on five trial runs (328, 310, 300, 290 and 280 documents), as it produced a clear five-cluster structure (119, 73, 43, 29 and 16 documents) while avoiding excessive fragmentation. This technique is widely applied in bibliometric reviews (e.g. Hallinger and Kovačević, 2019). The resulting cluster maps and text data were exported to Excel. Second, the data were sorted by clusters (ascending), citation counts (descending) and total link strength (descending). The authors then reviewed the documents by cluster, prioritizing highly cited studies and those with strong link strength. Each document was coded based on theoretical foundations and main findings, from which themes, research gaps and future research directions were inductively derived. Third, the two authors cross-checked their coding and any discrepancies were resolved through discussion and consensus.

To identify topical trends, we analyzed keyword co-occurrence, as it reflects conceptual relationships and emerging research trends (Hallinger and Kovačević, 2019). First, the dataset was standardized by merging synonymous keywords (e.g. digital HRM, DHRM and digital human resource management). Next, the data were analyzed in VOSviewer using the keyword co-occurrence technique. A minimum threshold of three occurrences per keyword was applied to ensure conceptual relevance while maintaining thematic diversity. The results included a keyword co-occurrence map and text data, which were exported to Excel and sorted by average publication year, cluster and total link strength. Finally, the two authors independently reviewed the literature associated with emerging keywords, identified contemporary themes and then conducted focused group discussions to reach consensus. Figure 2 summarizes the research design of the study.

Figure 3 illustrates the development trajectory of DHRM research from the first study in 2004 up to May 2025.

DHRM research developed slowly from 2004 to 2008, increased in 2009 and grew unevenly until 2018 as the field matured. Since 2019, publications have risen steadily. This trend reflects global disruptions, including the COVID-19 pandemic, which accelerated remote work and the adoption of digital HR systems.

Regarding the performance of individual countries, the literature shows that a total of 74 countries have contributed to the field of DHRM. Table 2 presents the countries with the highest contributions, measured by the number of documents published and the number of citations received.

European studies have a higher impact despite a lower volume, reflecting longer research traditions and higher-impact journals. Countries such as India and China have demonstrated strong recent publication growth, likely reflecting increasing investment in digital transformation within emerging economies. Differences in institutional and regulatory contexts, particularly regarding data privacy, algorithmic accountability and digital infrastructure readiness, may lead to significantly different patterns of DHRM adoption and governance. These findings highlight the need for more cross-national comparative research.

The bibliographic coupling analysis in VOSViewer identified five key themes, with the number of documents for each theme ranked from first to fifth as 119, 73, 43, 29 and 16, respectively. The results reveal the intellectual structure of DHRM, with five thematic clusters reflecting its developmental trajectory, from conceptual formation to identifying strategic roles, assessing impacts, analyzing influencing factors and ultimately advancing toward innovation. Figure 4 illustrates the science map based on these bibliographic coupling results.

This cluster captures the formative stage of DHRM research, when scholars focused on defining the field's conceptual boundaries and anchoring it within established theories. Early studies distinguished between transactional e-HRM and more strategic digital HR (Bondarouk et al., 2009), while later work framed e-HRM as an interplay of technology, people and processes (Strohmeier and Kabst, 2014) and emphasized the growing role of AI and predictive analytics (Bondarouk and Brewster, 2016). The cluster draws heavily on the resource-based view, technology acceptance model and dynamic capabilities to explain how digital HR capabilities create the competitive advantage.

This cluster reflects a dual orientation. Scholars seek to legitimize DHRM through established theories while debating whether it represents simple digitalization or strategic transformation. Theoretical diversity also indicates fragmentation, as an integrated framework linking technological, organizational and human dimensions remains absent.

This cluster directly engages with the debate on whether digitalization enables HR to move from administrative efficiency to strategic relevance. Research consistently shows that technology adoption alone does not guarantee strategic value; outcomes depend on implementation quality, managerial support and HR's analytical capabilities (Marler and Parry, 2016). While DHRM can streamline processes and support data-driven decision-making, its strategic contribution is contingent on organizational readiness and HR professionals' digital competencies (Bondarouk et al., 2017).

Analytically, this cluster marks a shift from aspirational claims to evidence-based assessment. The emphasis on contingency factors signals that the field has moved beyond technological determinism toward a socio-technical perspective, a finding that resonates with the key debates in Section 2.2. The persistent tension between efficiency gains and strategic value highlights a foundational question that remains unresolved, indicating an important area for future research.

This cluster views DHRM as a driver of organizational capability development and operational transformation. Digital HR practices strengthen learning agility and flexibility, while smart technologies such as IoT reshape how HR activities are coordinated and monitored (Strohmeier, 2020). Digital systems also enhance communication and integration across HR functions, though such benefits require strong alignment between business strategy and HR digitalization efforts (Schalk et al., 2013).

Research attention has expanded from individual outcomes to broader organizational capabilities. By linking DHRM to agility and dynamic capabilities, it suggests that digital transformation can drive bigger organizational change. However, the literature remains biased toward positive outcomes, with limited attention to how DHRM may simultaneously enable and constrain capability development.

This cluster focuses on the mechanisms through which DHRM may produce its effects, moving beyond simple input-output models. Scholars highlight improved data quality, stakeholder participation and service delivery as potential benefits (Bondarouk and Brewster, 2016), while also identifying boundary conditions such as organizational support and technological readiness. A notable feature is its engagement with the dark side of digitalization: some studies warn of reduced social interaction, accountability tensions and perceptions of dehumanization (Bondarouk and Brewster, 2016), and a sociomaterial perspective has been proposed to suggest that HR outcomes arise from the intertwined agency of humans and technologies (Myllymäki, 2021).

This cluster reflects a shift from direct-effect models toward understanding the conditions under which DHRM is effective. Greater attention to ethical concerns and algorithmic governance indicates a growing recognition of digitalization's unintended consequences. The adoption of a sociomaterial perspective further reinforces the field's broader move away from technological determinism.

This cluster reflects the most recent wave of research, focusing on how digital tools enable innovation across HR functions. Studies show that digital HR tools enhance service quality, productivity and decision-making and that system and information quality are central to the success of intelligent e-HRM platforms (McDonald et al., 2017). Digital HR also strengthens the impact of high-performance work practices and requires alignment between technological capabilities and HR strategies, especially in SMEs (L’Écuyer and Raymond, 2023).

Analytically, this cluster marks the field's shift toward forward-looking, innovation-oriented research. The prominence of AI and related technologies reinforces the emerging trends identified in Section 4.3. Its focus on SMEs and implementation challenges also highlights growing awareness of contextual differences in DHRM adoption.

Figure 5 presents a scientific map based on a co-occurrence analysis of keywords, with a minimum threshold of three occurrences per keyword. Brighter yellow colors indicate topics with a more recent average publication year (Hallinger and Kovačević, 2019). Some notable contemporary trends include AI, remote work, employee engagement and employee performance.

The first topical trend is AI's impact on HRM, attracting growing scholarly attention. AI enhances efficiency by automating recruitment, appraisal and training, improving objectivity and optimizing HR outcomes. It also aids strategic decisions via predictive analytics, real-time feedback and data insights, supporting candidate selection, turnover management and workforce planning (Zawada, 2024). However, AI raises ethical and operational challenges, including concerns about data privacy, bias and job displacement, thereby requiring transparency and human oversight. AI thus shifts HR from administrative tasks to strategic talent development (Zawada, 2024).

The second topical trend is remote work, which has become a lasting focus in HR and organizational management, especially after the COVID-19 pandemic (Orhan, 2024). Its rise reflects a sustainable shift driven by technology, social changes and employee demand for flexibility. Digital platforms enable work from anywhere, while the pandemic accelerated adoption as a survival strategy. Remote work offers benefits such as better work-life balance, flexibility, autonomy and higher engagement. Hybrid models are likely to persist, requiring ongoing adaptation and technological support (Orhan, 2024).

The third topical trend is employee engagement and performance, driven by digital transformation, which enhances communication, collaboration and work efficiency (Stachová et al., 2024). Remote work, accelerated by COVID-19, further underscores the need for technological engagement to boost organizational performance. Challenges remain, such as employee resistance and fear of job loss, requiring effective leadership and clear communication to maintain innovation.

The emergence of AI as a prominent keyword signals a shift in DHRM from process automation to intelligent decision support, closely linked to innovation and debates on algorithmic management (Theme 5). The rise of remote work following the COVID-19 pandemic further indicates DHRM's adaptation to new work models, challenging traditional HR systems. At the same time, employee engagement and performance have become critical concerns as organizations seek to sustain workforce effectiveness in increasingly digital and remote environments.

The future research directions proposed below are derived directly from the gaps identified in the thematic analysis and the emerging trends. Specifically, the analysis revealed that while the field has established a conceptual foundation (Theme 1) and recognized the strategic potential of DHRM (Theme 2), several areas remain underexplored: the mechanisms through which DHRM produces both positive and negative outcomes (Theme 4), the contextual factors shaping implementation success (themes 3 and 5) and the ethical tensions accompanying AI adoption (Section 4.3).

Thematic analysis shows that although Theme 4 addresses the dark side of digitalization, such as reduced social interaction, accountability tensions and perceived dehumanization, most studies remain conceptual rather than empirical. Similarly, although AI has become a dominant research focus (Section 4.3), its ethical implications are often acknowledged but are rarely examined systematically.

To address this gap, future research should employ experimental designs to examine how digitalization and automation affect employee perceptions of ethics, privacy and fairness. What theoretical frameworks can effectively explain ethical, privacy and fairness issues in DHRM in the AI era? What ethical principles should guide AI-driven recruitment, performance evaluation and employee monitoring to balance efficiency with fairness? How do employees perceive and respond to AI-based monitoring under different conditions of transparency and explainability?

The findings indicate a strong focus on positive outcomes, such as improved efficiency, service quality and decision-making (Theme 5), while negative consequences remain underexplored. Although Theme 4 highlights risks like dehumanization and reduced social interaction, empirical evidence on these impacts is still limited.

Future research should investigate the following questions: What are the negative impacts of digitalization and automation on employees and HRM practices? What are the consequences of excessive digitalization in HRM? How do emerging technologies, such as AI and humanoid robots, reshape HRM and the capabilities required of HR professionals? What are the limitations of AI in delivering empathetic feedback, personalized employee support and nuanced performance appraisal? How does prolonged interaction with AI systems affect employee well-being and perceptions of surveillance?.

A persistent tension identified across themes 2 and 3 concerns the relationship between technology and organization: while technology can restructure processes, organizational culture, digital literacy and human agency shape how technology is deployed. However, the literature has not yet clarified how AI can complement rather than replace human judgment in HR functions.

Future research should address the following questions: Which HR tasks are best suited for AI, and which require human judgment, empathy and contextual understanding? How can AI-generated insights be effectively integrated with managerial discretion in promotion and performance evaluation decisions? As automation and humanoid robots increasingly replace human labor, how should HRM strategies evolve and what organizational capabilities are needed to adapt?.

The topical trends analysis (Section 4.3) highlights the rapid adoption of AI and remote work, yet the literature remains focused on short-term outcomes such as process optimization and productivity gains (Theme 5). What remain underexplored are the long-term consequences of DHRM for individuals and organizations.

Future research should investigate how sustained exposure to AI-driven HR systems affects employee motivation, engagement and career development over time. It should also examine the implications of DHRM for organizational structure, labor relations and workforce adaptability in the face of continuous technological change. Longitudinal studies are particularly needed to identify factors that enable sustainable implementation rather than yielding short-lived efficiency gains.

The country-level analysis highlights the need to expand research in underexplored regions, including Latin America and Africa, as well as in cross-national or regional studies. Thematic analysis shows that DHRM research mainly targets organizations broadly, with small firms and SMEs underrepresented. Contextual factors, such as organizational support and technological readiness, are crucial for outcomes but remain insufficiently studied across diverse settings.

Future research should address the following questions: How do DHRM practices and outcomes vary across industries and regions with different cultural, legal and technological contexts? What are the key drivers and barriers to DHRM adoption in SMEs? Which factors enable SMEs to overcome financial, technological and capability constraints in implementing DHRM? Comparative and configurational approaches, such as QCA or multi-case designs, may help identify context-dependent pathways to successful implementation.

Across multiple themes, especially Theme 2 (strategic role) and Theme 3 (organizational capabilities), the findings indicate that technology alone does not determine outcomes; organizational culture, digital literacy and leadership are key mediators. However, systematic research on the interaction between digital culture and DHRM effectiveness remains limited.

Future research should explore the following questions: What are the benefits and limitations of a digital organizational culture? How should digital culture be developed in terms of strategic orientation and organizational priorities? What types of organizational culture are most conducive to successful digital transformation in HR? How does digital culture influence the implementation and effectiveness of AI systems in HR and what role does digital leadership play in supporting employees through technological change?.

This study contributes to DHRM theory in three ways. First, it provides a comprehensive overview of the field's evolution (2004–2025) and country-level performance gaps. It also highlights country- and region-level contribution performance, revealing contextual gaps. Second, the identification of five thematic clusters illustrates a developmental pathway: from foundational definitions (Theme 1) to strategic role assessment (Theme 2), organizational capability transformation (Theme 3), investigation of mediating mechanisms (Theme 4) and finally innovation-oriented research (Theme 5). These themes reveal a conceptual progression in DHRM, from technology adoption and process efficiency toward strategic transformation, ethical governance and socio-technical integration. This perspective extends prior reviews by highlighting how the field's intellectual focus has evolved. This theoretical framing may offer a conceptual scaffold to guide future theory integration. Finally, the study outlines research directions in the fields that are linked to emerging technologies such as AI, excessive digitalization and humanoid robotics, emphasizing the need to consider their dark sides and the strategies for balancing human-technology interaction in DHRM.

For practitioners, the findings suggest several directions worth considering. In terms of strategic orientation (Theme 2), realizing the potential of DHRM may require not only the adoption of technology but also digital competency readiness and alignment with business strategy. Regarding operations (themes 3 and 4), DHRM has the potential to enhance organizational agility, but managers should be mindful of dark side risks, such as reduced social interaction or perceptions of dehumanization; designing transparent systems and maintaining human oversight may be advisable. In the context of innovation (Theme 5), institutional and industry-specific factors appear to influence DHRM outcomes, suggesting that context-sensitive approaches may be more appropriate than a one-size-fits-all solution. Finally, for ethical AI governance, organizations may consider implementing bias-checking procedures, ensuring transparency and establishing feedback mechanisms for employees, as trust in DHRM may depend on perceived fairness and human accountability.

In conclusion, this study reviews DHRM research from 2004 to 2025, highlighting its growth trajectory, key contributing countries and five major thematic clusters. Three emerging trends were identified: AI, remote work and employee engagement. Based on the identified gaps, six future research directions related to emerging technologies are proposed. This study contributes an up-to-date synthesis and reveals a conceptual shift in DHRM toward strategic and socio-technical perspectives, offering a forward-looking agenda for technology-driven HR research. Limitations of this study include some bibliometric results are not presented, and other systematic review techniques and framework-based analyses are also not explored. Relying solely on Scopus and applying exclusion criteria may have omitted relevant publications. Future research should therefore include additional databases, broaden inclusion criteria and adopt more interdisciplinary approaches.

This article contains no studies with human participants performed by any author.

During the preparation of this work, the authors used ChatGPT and DeepSeek to improve the language and readability of the manuscript. After using these tools, the authors reviewed and edited the content as needed and accept responsibility for the content of the publication.