Abstract
Keywords
Introduction
The Fifth Industrial Revolution (5IR) is a period where human intelligence and advanced technologies such as artificial intelligence, robotics, and machine learning work together to support meaningful and ethical service delivery. Unlike the Fourth Industrial Revolution, which focused heavily on automation, 5IR places greater attention on human values, inclusivity, and responsible use of technology. In this evolving context, libraries are experiencing changes in how information services are delivered. The idea of the intelligent library reflects this shift, as AI tools are increasingly used to support decision making, improve user engagement, assist routine operations, and enable information professionals to provide personalized services.1,2 These developments require librarians to acquire practical AI-related skills and competencies that align with the expectations of 5IR service environments.
AI is now part of daily library operations rather than a distant innovation. Research shows that library leaders and practitioners recognize its potential to improve efficiency and strengthen service outcomes, though levels of readiness and skill development differ across regions. 3 AI applications are seen in cataloguing automation, discovery tools, natural language search, chatbots, recommendation systems, and data-driven services. 2 These applications require librarians to develop skills in data literacy, algorithmic thinking, and ethical use of digital tools, as well as broader competencies such as critical thinking, problem solving, and adaptability. As Cox et al. 1 explain, the future success of academic libraries depends not only on the presence of intelligent systems but also on the ability of librarians to manage and use them responsibly.
Although interest in AI is growing, questions remain about how prepared librarians are to work effectively in 5IR environments. Studies from different regions reveal varying levels of confidence and experience with AI tools. For example, Yoon et al. 4 report that librarians in North America appreciate the benefits of AI but express concerns about job displacement, ethical issues, and implementation challenges. In Africa, Lulu-Pokubo and Okwu 5 found that librarians recognize the value of AI for sustainable library services but face constraints such as inadequate technical knowledge, limited infrastructure, and policy gaps. These findings show the need for research that focuses on the specific skills and competencies librarians must possess to adopt AI effectively in 5IR libraries.
AI skills in librarianship refer to the technical abilities required to operate and interact with AI systems, including knowledge of machine learning concepts, the ability to work with data, an understanding of algorithmic processes, and familiarity with AI-driven tools. Competencies extend further to include ethical reasoning, user-centred decision making, professional adaptability, and the capability to integrate intelligent systems into service delivery. 5IR promotes responsible, inclusive, and sustainable use of technology, which means librarians must be able to address issues such as data privacy, fairness, and equitable access. Harisanty et al. 3 note that awareness alone is not enough; librarians must be able to convert that awareness into functional abilities before AI can be effectively integrated into library operations. Understanding whether librarians possess these abilities is essential for developing training programs, policy direction, and institutional support for AI adoption.
Across many regions, libraries are at different stages of adopting AI. Some institutions have incorporated advanced intelligent systems, while others are still struggling with basic technological requirements. This uneven progress makes it necessary to evaluate the actual skills and competencies of librarians who serve as key actors in this transition. By examining their capabilities, this study seeks to identify gaps and opportunities that can support professional development in 5IR libraries. The study positions librarians as central contributors who must apply both technical and ethical knowledge to ensure successful AI-enabled service delivery.
The integration of artificial intelligence in libraries is reshaping service delivery in the 5IR era where effective collaboration between human expertise and intelligent systems is essential. Despite the rapid adoption of AI tools, it is unclear whether librarians possess the skills and competencies required to use these technologies responsibly and efficiently. Existing studies show that librarians are aware of the benefits of AI3,5 and acknowledge its usefulness in improving service quality. 4 However, research also reports gaps in technical ability, ethical competence, and institutional support.1,2
Most previous studies focus on awareness, perceptions, or general attitudes toward AI in libraries. Very few examine the specific AI-related skills and competencies that librarians need for effective practice within the 5IR context. This creates a gap in understanding how prepared librarians are to move from mere awareness to practical implementation. This study addresses that gap by examining the AI skills and competencies required of librarians in 5IR libraries. The study is timely because AI adoption is increasingly expanding, and libraries risk losing relevance if professionals cannot apply these tools in ways that are ethical, inclusive, and aligned with 5IR principles.
Objectives of the study
The main objective of this study is to examine artificial intelligence (AI) skills and competencies among librarians in 5IR libraries. Specifically, the study seeks to: (1) Determine librarians’ knowledge of AI technologies relevant to library services in the 5IR era. (2) Identify the skills librarians possess for implementing AI applications in library operations and service delivery. (3) Examine the competencies required for AI adoption in 5IR libraries. (4) Explore the challenges librarians face in acquiring and applying AI skills and competencies.
Methodology
Research design
This study adopted a systematic literature review approach guided by the Preferred Reporting Items for Systematic Reviews and Meta Analyses (PRISMA) framework. The study followed the systematic literature review method because it enables the synthesis of scattered evidence and supports a clear understanding of current issues in library and information science, as seen in the works of Lee et al. 6 and Mabon et al. 7 The PRISMA model provides a structured and transparent method for identifying, screening, and evaluating research evidence. PRISMA guided the stages of identification, screening, eligibility assessment, and inclusion. This structured approach strengthened the integrity and transparency of the review and supported a focused examination of studies that address artificial intelligence related skills and competencies among librarians in 5IR libraries.
Selection of information resource databases
Studies retrieved across databases.
Search terms and search strategy
Search terms were developed to reflect the central concepts of artificial intelligence, skills, competencies, librarians, and 5IR libraries. The strings applied in Scopus, LISA, and Google Scholar included TITLE ABS KEY (“artificial intelligence” OR “AI technologies” OR “intelligent systems” OR “machine learning” OR “robotics”) AND (“librarian*” OR “library staff” OR “information professional*”) AND (“skills” OR “competenc*” OR “expertise” OR “readiness” OR “training”). The Web of Science search adopted the expression TS = (“artificial intelligence” OR “AI technologies” OR “intelligent systems” OR “machine learning” OR “robotics”) AND (“librarian*” OR “library staff” OR “information professional*”) AND (“skills” OR “competenc*” OR “expertise” OR “readiness” OR “training”). These search terms ensured retrieval of studies that examine not only awareness of AI but also the practical skills and professional capabilities needed for AI adoption in 5IR libraries. The search was carried out between July 2025 and August 2025.
Inclusion and exclusion criteria
Studies were included if they were published between 2019 and 2025, written in English, and classified as journal articles or conference papers. They were also required to focus directly on AI-related skills, competencies, readiness, or training needs of librarians within library settings. Publications were excluded if they concentrated on AI in unrelated fields such as general education or computer science, mentioned AI without discussing librarian competencies, lacked full-text access, or failed to meet acceptable standards during quality appraisal. These conditions ensured that the review incorporated only studies that were credible, relevant, and aligned with the purpose of examining AI competencies in 5IR libraries.
Data collection and screening
The initial search across the four databases produced 457 records. These citations were exported into Microsoft Excel for cleaning, where duplicates and triplicates were removed, resulting in 328 unique publications. Screening involved evaluating titles and abstracts to determine relevance to the study objectives, and publications that did not address AI competencies in librarianship were removed. After this stage, 146 articles qualified for full-text review. The eligibility process involved carefully reading each article to confirm alignment with the focus on AI skills, competencies, challenges, or training needs among librarians. During this review, 86 articles were excluded for lack of relevance and 12 could not be accessed in full text.
Quality appraisal
All eligible articles were appraised using the Liu and Liu 8 checklist, which examines methodological soundness, clarity of reporting, and relevance of findings. Only studies that met the minimum quality requirements were included. Any differences in judgment between the researchers were resolved through discussion to ensure consistent application of the criteria. A total of 48 articles met all requirements and were included in the final synthesis. The flow of identification, screening, eligibility assessment, and inclusion was documented using the PRISMA diagram recommended by Faulkner and Reiter. 9
Data analysis
The final set of articles was subjected to narrative analysis. This involved extracting information related to study objectives, definitions of AI skills and competencies, described training needs, methodological features, and reported challenges. The extracted content was organized and interpreted to identify recurring ideas, differences across regions, and emerging patterns. Narrative analysis was appropriate because it allowed the study to integrate diverse findings and present a coherent explanation of the evidence on AI skills and competencies among librarians in 5IR libraries (Figure 1). Procedure for inclusion and exclusion.
Details of selected publication included in the study.
Results and discussion
Librarians’ knowledge of AI technologies relevant to library services in the 5IR era
Studies show that librarians generally recognize AI as useful for core services such as cataloguing, discovery, chatbots, recommendation systems, and automation. Research in Nigeria reports broad awareness of AI’s potential, although this awareness often remains conceptual.10,11 Similar findings from North America indicate that librarians accept AI as relevant but differ in how they define it. 12 More recent studies confirm that while librarians increasingly view AI as important for emerging services, their understanding is still more conceptual than technical.13,14
Although awareness is widespread, technical knowledge remains limited. Studies report that librarians can identify AI powered tools such as chatbots, automated metadata enrichment systems, recommender engines, and analytics dashboards but do not fully understand how these tools work or how underlying models are trained and evaluated.12,15 This means that many librarians can describe AI applications yet depend on technical support when assessing implementation requirements. Findings also show variation in practical knowledge. Some librarians have working familiarity with AI tools such as APIs, basic chatbot setup, or analytics use, but a larger number have only surface-level familiarity or no hands-on experience.11,16 Studies from developing regions such as Nigeria, Ghana, Zambia, and Pakistan link this gap to limited resources and restricted access to professional development.11,16,17 Research also reports considerable variation in how librarians define AI. Hervieux and Wheatley observed disagreement in definitions, and other studies show that some librarians view AI as general automation while others associate it with machine learning, natural language processing, or neural models.12,18
Limited formal education and continuing professional development contribute to these knowledge gaps. Studies on curricula and competencies show that many LIS programmes and in-service training still emphasize traditional subjects and rarely include AI, data science, or analytics modules.13,15,19 Regional studies also note that librarians often obtain AI knowledge informally through workshops or self-study.20,21
Knowledge of ethics, privacy, and governance related to AI is developing but uneven. Studies report that librarians are concerned about issues such as bias, privacy, and intellectual property, although structured knowledge of fairness, data governance, or legal implications remains inconsistent.12,13,22 Librarians tend to feel more prepared to discuss user-facing concerns than to assess technical fairness or audit AI models. Finally, recent studies also show growing interest in AI training, pilot projects such as chatbots and discovery services, and institutional strategies that support upskilling.11,21–23 These developments suggest improving knowledge in some settings, although most authors agree that structured curriculum reform, coordinated continuing professional development, and practical project-based learning are needed to strengthen librarians’ technical understanding in the 5IR era.
Skills librarians possess for implementing AI in libraries
Studies show that librarians generally possess foundational awareness of AI and basic digital literacy, although this rarely develops into advanced technical skills. Research in Nigeria and similar contexts reports that librarians recognize AI’s potential for cataloguing, discovery, and user services, but most knowledge remains introductory and does not translate into practical mastery of AI tools.10,11,20 These studies show that librarians can describe expected functions of AI, such as automation and enhanced search, yet many respondents report only basic experience with digital tools and minimal hands-on work with AI systems.
The review indicates that librarians consistently demonstrate non-programming competencies that support AI adoption. These include information literacy, metadata and classification skills, user-centred service design, research methods, and critical evaluation capabilities. These competencies are viewed as valuable for preparing datasets, curating training corpora, interpreting AI outputs, and designing user workflows.15,24,25
Technical skills remain limited across many regions. Evidence from empirical studies on big data and AI competencies shows restricted proficiency in programming languages, data cleaning, statistics, and machine learning concepts. Where such skills exist, they tend to be concentrated among librarians with additional IT or data science training.15,26,27 In addition, ethical and AI literacy are increasingly recognized as necessary skills. Studies indicate that librarians are building competencies related to bias, privacy, transparency, and regulatory awareness, and many show interest in providing user education on AI-related issues.20,23 While librarians express confidence in addressing general ethical concerns, they also identify the need for formal training to strengthen their knowledge.12,13,24 This positions librarians as intermediaries between AI systems and users even when they are not involved in programming.
Evidence from Nigeria and neighbouring African countries shows that curriculum limitations and institutional constraints continue to affect skills development. Authors report that many LIS programmes still prioritize traditional subjects and offer limited AI-focused content. The scarcity of continuing professional development opportunities and inadequate institutional support also contribute to limited technical capacity for coding, data science, and system integration.11,17,20 Recent studies show that librarians’ AI confidence improves when targeted training is provided, demonstrating that skills gaps can be addressed through structured programmes.20,24
Competencies required for AI adoption in 5IR libraries
The review shows that effective AI adoption in 5IR libraries requires a combination of technical, cognitive, and ethical competencies that extend beyond traditional practice. Several studies state that librarians need AI literacy to work meaningfully with AI tools and vendors, interpret algorithmic decisions, recognize bias, and maintain transparency in service delivery.13,19,28 These competencies are essential for sustaining equity and intellectual freedom in AI supported environments,22,24 making AI literacy a core requirement on which other skills depend.
Technical competencies remain central across the studies. Research identifies skills in data management, big data analytics, data cleaning, and introductory programming in languages such as Python and R as important for managing training data and interacting with machine learning systems.15,26 Other works emphasize computational thinking, database design, and system integration for deploying chatbots, recommender tools, and predictive analytics.11,16,27 Competence in technology evaluation and procurement is also viewed as necessary, since many libraries will acquire external AI tools rather than build them. 29
Ethical and governance competencies form a major theme in the literature. Studies emphasize the need for knowledge of privacy regulations, copyright considerations, and principles of fairness, accountability, and transparency when using AI systems.12,30–32 These skills support the creation of responsible AI policies and user education. Work in school and public libraries also shows that librarians increasingly serve as AI literacy educators who help learners engage critically with AI tools.33–37
Strategic and managerial competencies are frequently identified as important for leading AI adoption. Studies recommend skills in strategic planning, project management, and change management to support funding, staffing, and sustainable implementation.20,23 Collaborative leadership and teamwork are necessary because AI projects often involve IT professionals, data scientists, and academic partners.25,38 Affective competencies also contribute to adoption. Librarians benefit from openness to innovation, problem solving ability, and adaptability as they work with evolving tools and address concerns about job insecurity.20,39–42 Communication skills also support the ability to explain AI benefits and reduce user anxiety.20,41
Continuous professional development emerges as an essential competency across regions. Studies recommend ongoing upskilling, certifications in data analytics or AI, and inclusion of AI content in LIS curricula to prepare librarians for evolving technological expectations.32,43 This need for continuous learning is echoed in a wide range of contexts, indicating the importance of adaptable skills for supporting AI in 5IR libraries.40,44
Challenges of implementing AI in library
The review identifies one of the major challenges facing librarians in implementing AI is tied to limited formal education and gaps in LIS curricula, leaving many librarians without structured pathways to learn AI-related knowledge and practical skills. Studies report that many programmes still prioritize traditional subjects and offer little sustained training in data science, machine learning, or applied AI, which forces librarians to depend on short workshops or self-directed learning.11,15,19,45
Insufficient continuing professional development also contributes to skill gaps. Research from Nigeria and comparable settings shows that although librarians express a strong interest in learning AI, institutional training opportunities are irregular, underfunded, or not aligned with practical tasks.11,20,46 Resource and infrastructure constraints present an added barrier. Studies from developing-country contexts describe limited computing power, unreliable internet, and the absence of test environments, which restrict opportunities to practise AI implementation and reinforce dependence on vendor platforms.17,23,47
Technical difficulties and perceptions of AI complexity are also widely reported. Literature shows that many librarians lack confidence in programming, statistics, and data engineering skills needed for practical work.15,26 Limited internal technical capacity often pushes libraries toward purchasing commercial solutions rather than developing tools in-house, which reduces hands-on learning.29,48 Institutional and managerial barriers further affect adoption. Studies show that limited leadership support, unclear strategies, inadequate budgets, and a lack of collaboration with IT units hinder progress, making AI initiatives difficult to sustain when responsibilities are not clearly assigned.20,23,32,49–55
Ethical and legal uncertainties also influence librarians’ willingness to apply AI skills. Concerns about data privacy regulations, copyright issues, model bias, and unclear procurement agreements make librarians hesitant to deploy AI systems without adequate policy guidance.12,13,22 Several studies report that apprehension about automation, fear of job loss, and anxiety about rapid technological change reduce engagement with training and experimentation.10,39,56 Research indicates that confidence improves when training includes clear communication about role evolution, supportive leadership, and opportunities for inclusive participation.
Conclusion
This review set out to examine artificial intelligence (AI) skills and competencies among librarians within the context of the Fifth Industrial Revolution. The systematic review of 48 peer-reviewed studies revealed that while librarians generally possess a good conceptual understanding of AI and acknowledge its potential for transforming library services, there is a clear gap between awareness and technical application. Most librarians are familiar with AI applications such as chatbots, recommender systems, automated metadata enrichment, and analytics dashboards, but they often lack in-depth knowledge of the underlying mechanisms that power these systems, including machine learning models, natural language processing, and data management workflows. This highlights the need for structured learning pathways to build practical, hands-on competence rather than relying solely on theoretical or vendor-driven exposure.
The study also established that librarians possess strong information literacy, metadata expertise, and user-centred service design skills. However, technical proficiencies such as programming, data cleaning, algorithmic evaluation, and system integration remain underdeveloped in most regions, especially in developing countries. These gaps make libraries dependent on external vendors and limit their capacity to independently design, implement, or critically evaluate AI solutions. Ethical reasoning, data governance literacy, and the ability to teach AI literacy to patrons are emerging areas where librarians are beginning to develop competence, but there is still a need for formal training to strengthen these capacities in line with 5IR expectations.
Furthermore, the study identified several challenges that hinder librarians’ ability to acquire and apply AI skills. These include outdated LIS curricula that do not embed AI literacy, insufficient continuing professional development opportunities, inadequate infrastructure and resources, and a lack of institutional strategies and leadership support for AI adoption. Ethical and legal uncertainties, staff anxiety about job displacement, and the complexity of AI tools further slow down skill uptake. Collectively, these barriers implied the importance of a coordinated, systemic response that combines curriculum reform, institutional policy development, strategic investment in training and infrastructure, and deliberate change management initiatives. As a result, the findings suggest that for libraries to remain relevant and competitive in the 5IR era, librarians must be supported to transition from conceptual awareness to practical mastery of AI technologies. This will require a multi-pronged approach involving curriculum redesign to include AI and data science modules, sustained professional development programmes, cross-disciplinary collaborations with IT and data science experts, and the creation of enabling institutional environments that prioritize innovation, ethical governance, and inclusive skill-building.
The findings of this study have significant implications for both professional practice and policy formulation in academic librarianship as well as LIS education/curricula. At the practical level, the evidence indicates the urgent need for academic libraries to invest in structured and continuous capacity-building programmes that equip librarians with both foundational AI literacy and advanced technical competencies. Practical training modules should move beyond theoretical introductions to AI and focus on hands-on experience with tools such as machine learning-based cataloguing systems, recommender engines, and data analytics platforms. From a policy perspective, the findings call for deliberate integration of AI and data science content into Library and Information Science (LIS) curricula at undergraduate and postgraduate levels. National professional bodies and accreditation agencies should update curricular standards to reflect the skills and competencies required for the 5IR era, ensuring that graduates are industry-ready and capable of engaging with emerging technologies. Additionally, Library and Information Science (LIS) education requires strategic reform to prepare future professionals for the growing use of artificial intelligence in library services. Many of the studies reviewed indicate that current LIS curricula still emphasize traditional areas such as cataloguing, reference services, and collection management, while providing limited coverage of AI tools, data literacy, and automated decision systems.15,19 One priority is the inclusion of structured modules on AI concepts, machine learning basics, natural language processing, and practical exposure to AI-enabled library systems. These modules should introduce students to how intelligent systems work and how they can be applied responsibly within library settings. Alongside technical skills, LIS schools should embed data ethics, algorithmic fairness, privacy regulation, and responsible automation as core elements of the curriculum. In addition, digital skills training needs to be expanded to cover data literacy, data management, basic programming, and the ability to evaluate automated outputs. Practical laboratory sessions, collaborations with computer science departments, and project-based learning can help students build confidence in real AI applications. Professional competencies such as change management, technology evaluation, and cross-disciplinary teamwork should also receive attention, since they are essential for librarians who will guide AI adoption in institutions. Additionally, accreditation bodies, national LIS associations, and university quality assurance units should develop guidelines that recognize AI literacy, data governance, and ethical digital practice as compulsory components of LIS training.
However, this study has some limitations. First, the review relied primarily on published peer-reviewed journal articles and conference proceedings, which may have excluded relevant insights from grey literature and professional reports. This may have narrowed the scope of perspectives included, particularly from practitioners working outside academic publishing circles. Second, while the study synthesized findings from multiple geographic regions, there is an overrepresentation of studies from Africa and Asia, which may limit the generalizability of the results to global academic librarianship contexts. Third, the review focused largely on qualitative and survey-based studies, and fewer empirical studies with experimental or longitudinal designs were available for inclusion. This limits the ability to infer causal relationships between AI skills acquisition and successful implementation in libraries.
Future research should seek to address these limitations by incorporating a wider range of literature, including grey sources and professional practice documents, to provide a more holistic understanding of AI skills and competencies. Comparative and longitudinal studies across diverse global contexts are also recommended to capture variations in AI adoption and skill development over time. Additionally, experimental studies or pilot interventions could be conducted to evaluate the effectiveness of targeted training programmes and institutional policies in building AI readiness among librarians.
Footnotes
Acknowledgements
ORCID iD
