AI-Related Curriculum Visibility in Selected ASEAN LIS Programs: An Exploratory Public-Document Mapping Study
Abstract
This exploratory study maps artificial intelligence (AI)-related curriculum visibility in selected ASEAN Library and Information Science (LIS) program records through public-document curriculum mapping, descriptive frequency analysis, and thematic coding. It is framed as an exploratory pilot study rather than a census or ranking of ASEAN LIS programs. Using 18 publicly available program records and 25 course-level or course-like curriculum signals from eight ASEAN countries, the study distinguishes explicit AI/ML integration from data-digital-analytics pathways and foundational ICT/information-systems preparation. It also compares visible curriculum signals with selected AI literacy and LIS professional competency frameworks, while accounting for uneven public-document availability, evidence confidence, and the limits of inferring enacted curriculum from web-accessible records. Findings suggest that AI-relevant preparation is more often visible through digital libraries, information retrieval, databases, analytics, data mining, and systems work than through explicitly AI-titled curriculum. The main contribution is a transparent curriculum-visibility and competency-alignment framework that can support later verified benchmarking, syllabus-level analysis, and regional curriculum development.
Keywords
artificial intelligence, library and information science education, curriculum mapping, competency-based education, ASEAN, Southeast Asia
1 Introduction
1.1 Artificial intelligence as a curriculum problem in LIS
Artificial intelligence (AI) has moved from a peripheral technology concern to a structural condition of contemporary information work. Search, discovery, metadata creation, scholarly communication, research support, reference work, digital preservation, bibliometrics, and information literacy instruction are increasingly shaped by algorithmic systems. For Library and Information Science (LIS) education, the issue is therefore not only whether students can use particular tools, but whether curricula prepare future professionals to understand, evaluate, govern, and teach with AI in ways consistent with librarianship’s commitments to access, intellectual freedom, privacy, equity, and public service.
Recent LIS scholarship frames this challenge as a curriculum alignment problem. A scoping review and syllabi analysis of AI literacy in LIS education found substantial fragmentation: some programs foreground technical and analytical competencies, while others emphasize critical, ethical, and sociotechnical interpretation, and many lack coherent learning objectives or assessment criteria for AI literacy (Bridges et al., 2025). This fragmentation matters because AI is no longer confined to specialist computing contexts. It is embedded in discovery systems, research platforms, content generation tools, analytics systems, and everyday user questions. In this context, AI literacy requires more than operational familiarity. It includes technical, ethical, critical, and societal dimensions (Lo, 2025).
Professional competency guidance has begun to articulate the same shift. The Association of College and Research Libraries (ACRL) identifies AI competencies for academic library workers across ethical considerations, knowledge and understanding, analysis and evaluation, and use and application (Association of College and Research Libraries, 2025). These domains are especially relevant for LIS curricula because they bridge practical tool use with professional judgment. The ACRL competencies position library workers not simply as AI users, but as evaluators of reliability, appropriateness, values alignment, and community impact. Internationally, IFLA’s statement on libraries and AI similarly signals that AI adoption must be assessed in relation to information rights, equity, privacy, literacy, automation, and governance (International Federation of Library Associations and Institutions, 2025).
1.2 The ASEAN LIS curriculum context
ASEAN is a useful regional setting for this question because LIS education, national quality assurance systems, digital infrastructure, and AI policy environments are not developing at the same pace. Earlier work on ASEAN LIS education identified quality assurance and regionalization as central concerns for the development of LIS programs (Sacchanand, 2015). More recently, AI has become part of the region’s broader digital policy agenda. The ASEAN Responsible AI Roadmap 2025–2030 calls for responsible AI to be operationalized across member states through governance, capacity building, infrastructure, and interoperability (ASEAN Secretariat, 2025). For LIS education, that regional policy direction raises a practical curriculum question: whether schools preparing information professionals are also making AI-related competencies visible in formal curriculum structures.
A 2025 study of AI adoption in ASEAN libraries shows why the curriculum question matters beyond the classroom. The study analyzed 115 official documents and found that AI adoption in ASEAN libraries is shaped by national strategic support, implementation capacity, and institutional challenges and opportunities (Y.-S. Xu et al., 2025). If libraries are experimenting with AI for discovery, user engagement, analytics, and operational efficiency, LIS curricula become one mechanism for converting regional AI ambition into professional capability.
The initial ASEAN curriculum harvest for this study suggests that AI integration is visible but uneven. Some programs name AI directly, as in the University of Southern Mindanao’s Artificial Intelligence (AI) Application in Libraries course (University of Southern Mindanao, 2024) and VNU-USSH Hanoi’s Digital Library and Artificial Intelligence specialization (University of Social Sciences and Humanities, Vietnam National University Hanoi, 2025). Other programs integrate AI-adjacent competencies through analytics, text and web mining, data extraction, data science, digital libraries, research data management, information retrieval, metadata, database systems, programming, and web technologies. NTU Singapore’s MSc Information Studies, for example, includes text and web mining, data extraction, social media analytics, user metrics, information visualization, and information mining and analysis (Nanyang Technological University, 2026). UiTM Malaysia, Chulalongkorn University, Universitas Indonesia, and Politeknik Brunei show different configurations of digital libraries, data analytics, digital humanities, database systems, information retrieval, and library informatics (Chulalongkorn University Department of Library Science, 2026; Politeknik Brunei, 2026; Universitas Indonesia Faculty of Humanities, 2025; Universiti Teknologi MARA, 2026).
| country | records | profile | signals |
|---|---|---|---|
| Brunei | 1 | Data/digital/analytics pathway (n=1) | Only higher institution in Brunei offering a Level 5 librarianship course; covers libraries, information centers… |
| Indonesia | 2 | AI-adjacent foundations (n=1); Data/digital/analytics pathway (n=1) | Metadata Informasi; Pangkalan Data untuk Lembaga Informasi; Humaniora Digital; Aplikasi Teknologi Pengelolaan In… |
| Malaysia | 3 | Data/digital/analytics pathway (n=2); AI-adjacent foundations (n=1) | ICT Application in Library and Information Science; Digital Reference Services; Web Development and Library Port… |
| Myanmar | 2 | AI-adjacent foundations (n=2) | Department offers undergraduate, diploma, master, MRes, and PhD programs; research areas include online library … |
| Philippines | 5 | Data/digital/analytics pathway (n=3); AI-adjacent foundations (n=1); Explicit AI/ML integration (n=1) | Programming Fundamentals; Database Design for Libraries; Information Processing and Handling in Libraries and In… |
| Singapore | 1 | Explicit AI/ML integration (n=1) | Digital Libraries; Social Media Analytics; Text and Web Mining; Data Extraction Techniques; User Metrics and Ana… |
| Thailand | 2 | AI-adjacent foundations (n=1); Data/digital/analytics pathway (n=1) | Information Retrieval System; Digitization; Web Design and Development in Information Work; Data Communications … |
| Vietnam | 2 | Data/digital/analytics pathway (n=1); Explicit AI/ML integration (n=1) | Applying AI tools in library operations and knowledge management; smart library services; digital information re… |
Table 1 positions the ASEAN evidence as a deliberately modest seed sample rather than a complete census of LIS schools. Its value at this stage is diagnostic: it shows that AI-related curriculum integration cannot be inferred from the presence or absence of one course title. The regional evidence includes explicit AI, machine learning, text/web mining, analytics, data mining, data science, digital libraries, information retrieval, metadata, databases, programming, and systems design. These curricular signals are not equivalent, but each represents a possible route through which AI-relevant competencies enter LIS education.
Figure 1 makes the same point visually. Within this seed sample, explicit AI/ML integration appears, but it is not the dominant pattern. More common are data, digital, and analytics pathways or foundational ICT pathways. This pattern is analytically important because curriculum mapping that searches only for the phrase “artificial intelligence” would undercount AI-relevant preparation, while a mapping that treats all digital courses as AI integration would overstate curricular responsiveness.
1.3 Research gap and study purpose
The global literature increasingly recognizes uneven AI integration in LIS curricula (Bridges et al., 2025), and professional bodies have begun to define AI literacy and AI competencies for library work (Association of College and Research Libraries, 2025; International Federation of Library Associations and Institutions, 2025). However, ASEAN LIS programs remain underexamined as a regional curriculum field. Existing ASEAN LIS education scholarship foregrounds quality assurance and regionalization (Sacchanand, 2015), while emerging ASEAN AI policy foregrounds responsible AI capacity and implementation (ASEAN Secretariat, 2025). What remains less visible is how LIS schools across the region translate emerging AI-related professional requirements into course titles, course descriptions, electives, specializations, and competency emphases.
This gap is consequential for three reasons. First, ASEAN LIS graduates are entering libraries, archives, schools, universities, government agencies, and information centers where AI-mediated systems are increasingly part of everyday information work. Second, AI-related curriculum integration may be uneven across countries and institutions, producing unequal preparation for graduates from different programs. Third, without a regional baseline map, curriculum revision risks becoming reactive: programs may add isolated AI content without clarifying whether students are expected to develop technical fluency, critical AI literacy, data competencies, ethical judgment, service design capacity, or governance-oriented professional practice.
This study addresses that gap by mapping visible AI-related curriculum signals in selected publicly available ASEAN LIS program records. The object of analysis is not enacted teaching, student competence, faculty expertise, or institutional quality. It is the curriculum evidence that institutions make publicly visible. The study asks:
- What explicit AI, machine learning, NLP, generative AI, data analytics, information retrieval, metadata, digital library, and governance-related curriculum signals are visible in publicly available ASEAN LIS program documents?
- How can the identified curriculum signals be classified by integration intensity: foundational ICT/information systems, data-digital-analytics pathway, or explicit AI/ML integration?
- Which AI-related professional competency domains are visibly supported by the curriculum signals when compared with selected AI literacy and LIS professional competency frameworks?
- What competency-alignment gaps are visible in the public curriculum evidence, particularly in relation to generative AI, algorithmic evaluation, bias, privacy, intellectual property, governance, and AI-supported user education?
- How does public-document availability affect the comparability of AI curriculum evidence across selected ASEAN LIS program records?
1.4 Conceptual framework: curriculum visibility and competency alignment
The study combines Curriculum Theory and Competency-Based Education (CBE) in a Curriculum Visibility–Competency Alignment Model. Curriculum Theory directs attention to what knowledge is selected, organized, sequenced, legitimized, and assessed within formal educational programs. It treats curriculum not as a neutral list of courses, but as a structured representation of professional priorities. In this study, that perspective supports close attention to whether AI appears as a required topic, an elective, a digital-systems proxy, a specialization, or a competency that is not visible in public documents.
CBE complements this lens by asking whether curricular content aligns with the competencies graduates need for professional practice. For AI in LIS, those competencies include not only technical understanding, but also ethical evaluation, responsible use, information literacy instruction, algorithmic critique, data governance, privacy awareness, and service design judgment (Association of College and Research Libraries, 2025; Lo, 2025). Together, Curriculum Theory and CBE make it possible to compare formal curriculum evidence with emerging professional requirements and to identify both visible content and consequential silences.
The model has five operational constructs. Curriculum visibility refers to whether AI-related content is publicly visible in formal program documents, such as course titles, course descriptions, specialization names, program outcomes, catalogues, and prospectuses. Integration intensity classifies visible signals as foundational ICT/information systems, data-digital-analytics pathways, or explicit AI/ML integration. Competency alignment compares the signal with professional domains such as technical fluency, data practice, algorithmic evaluation, ethical reasoning, privacy and governance judgment, AI-supported information literacy instruction, service design, and policy or procurement awareness. Curriculum gaps are claimed only as missing or weakly visible alignments relative to benchmark frameworks, not as proof that a program does not teach the area. Evidence confidence moderates every claim by distinguishing stronger course-level documents from thinner public program pages.
Figure 2 visualizes the study’s conceptual framework by showing how Curriculum Theory and CBE lead to the five operational constructs and to the study’s boundary conditions.
This framework therefore imposes two boundary conditions. First, public curriculum evidence can support claims about visible curriculum signals, not actual classroom delivery, assessment quality, faculty practice, or graduate competence. Second, digital libraries, databases, metadata, information retrieval, analytics, ICT, and AI are not conceptually equivalent. They are coded as related only when they plausibly provide foundations, data pathways, or explicit AI/ML preparation.
1.5 Contribution
This article makes three contributions. Empirically, it provides a structured exploratory map of visible AI-related curriculum signals in selected ASEAN LIS program records. Methodologically, it demonstrates a public-document curriculum-mapping approach that distinguishes explicit AI, data-digital-analytics pathways, and foundational ICT/information-systems preparation while recording evidence confidence. Practically, it offers a cautious competency-alignment framework for curriculum review and later benchmarking. By separating direct AI instruction from adjacent digital preparation, the study supports curriculum recommendations that are precise enough for program revision and bounded enough to avoid overclaiming from uneven public evidence.
3 Methodology
3.1 Research design
This study uses a qualitative-dominant content analysis design supported by descriptive frequency analysis. Content analysis is appropriate because the empirical material consists of public curriculum documents, program pages, course descriptions, and policy or competency documents that can be systematically unitized, coded, compared, and interpreted (Krippendorff, 2019). The design also uses document analysis, which treats documents as meaningful research data rather than neutral containers of facts. Following Bowen, documents are reviewed, extracted, coded, and interpreted to generate empirical understanding while retaining attention to context, provenance, and limitations (Bowen, 2009).
The study is exploratory and comparative only in the limited sense that it compares selected public program records using a common coding scheme. It does not claim to produce a complete census of all ASEAN LIS schools, a country ranking, or a measure of actual instructional quality. Instead, it builds a transparent seed dataset from publicly available evidence and uses curriculum mapping to identify patterns of visible AI-related integration. Curriculum mapping is suitable because it links curriculum content with expected competencies, learning outcomes, standards, or professional requirements (Arafeh, 2016). In this study, the map connects course and program evidence to the competency demands identified in the literature and professional guidance.
3.2 Data sources and sampling frame
The unit of institutional sampling is the LIS program record. A program record refers to a publicly documented LIS, information studies, library management, library informatics, or closely related degree/specialization offered by an ASEAN higher education institution. The initial harvested dataset contains 18 program records across eight ASEAN countries: Brunei, Indonesia, Malaysia, Myanmar, the Philippines, Singapore, Thailand, and Vietnam. Cambodia and Laos were searched but did not yield sufficiently clear public LIS curriculum records during the first two harvesting passes; this is treated as a data availability limitation rather than evidence of absence.
The evidence base includes three categories of documents. First, primary curriculum or program sources were harvested from university websites, academic catalogues, curriculum pages, prospectuses, and program PDFs. Second, regional or professional framework sources were gathered to define the competency-demand side of the study, including ASEAN AI policy, IFLA and ACRL AI guidance, and ASEAN LIS quality assurance literature. Third, selected scholarly studies were used to support the methodology and interpretive framework. The full source inventory and program-level dataset are stored in the project harvest directory and are treated as auditable research artifacts.
Program records were included only when a public source contained at least one curriculum-relevant item: an official curriculum plan, course list, course catalogue, program learning outcome, specialization description, prospectus, or handbook. Pages containing only general marketing claims were excluded from curriculum coding. The study also applies exclusion rules to the AI construct: generic ICT, databases, metadata, digital libraries, information retrieval, and analytics are not treated as equivalent to AI. They are coded as foundational or pathway signals unless AI, machine learning, NLP, text/web mining, generative AI, data mining, algorithmic evaluation, or AI governance is directly named or strongly implied by the course evidence.
3.3 Data harvesting procedure
Data harvesting was conducted through iterative web searching, source screening, and structured extraction. Searches combined country, institution, and discipline terms such as “Library and Information Science,” “Information Studies,” “Library Management,” “Library Informatics,” “curriculum,” “course,” “digital library,” “data analytics,” “artificial intelligence,” and “ASEAN.” Searches also used local institutional names and country-specific higher education terminology when discoverable. When available, primary institutional sources were preferred over secondary summaries. Secondary sources were retained only when they clarified regional context or provided historical/comparative evidence that could not be replaced immediately by a current primary document. Because multilingual source discovery was uneven across countries, language visibility is treated as a methodological constraint rather than a neutral feature of the data.
For each program record, the following fields were extracted: country, institution, program name, level, source status, source URL, access date or source year, AI integration score, AI integration label, AI-related curriculum signals, competency emphasis, thematic codes, preliminary gap note, and verification priority. A second course-level file extracts more granular evidence where course-level data were available, including course title, code, credits, required/elective status, AI relevance level, topic codes, evidence summary, and source URL. This two-level structure allows the study to distinguish broad program positioning from course-level curriculum evidence.
3.4 Units of analysis
The primary unit of analysis is the program record because public curriculum evidence is uneven across institutions. Some institutions publish detailed course descriptions, while others provide only program outcomes or broad course lists. The secondary unit of analysis is the course or course-like curriculum signal. A course-like signal includes a named course, specialization outcome, program outcome, or research/career competency statement that explicitly points to AI-related knowledge or skill.
This distinction reduces false precision. Where full syllabi are available, course-level coding can support stronger claims about curriculum depth. Where only program pages are available, coding remains at the program-signal level and is interpreted cautiously. In both cases, the analysis prioritizes transparent evidence over inferred institutional intent.
The third analytic unit is the topic or competency code. Topic codes identify the visible content domain, such as artificial intelligence, machine learning, generative AI, NLP/text mining, data mining, data analytics, information retrieval, metadata, digital libraries, research data management, privacy/data governance, AI ethics/bias/fairness, AI literacy/instruction, or human-centered service design. Competency codes interpret the professional capability implied by the visible signal, such as technical fluency, data practice, algorithmic evaluation, ethical reasoning, privacy/governance judgment, AI-supported information literacy instruction, service design and assessment, or policy/procurement awareness.
3.5 Coding scheme
The coding scheme was developed deductively from the research questions and competency literature, then refined inductively during source harvesting. The deductive frame distinguished explicit AI, AI-adjacent digital/data competencies, foundational ICT competencies, and professional/ethical competencies. Inductive refinement added recurring topic codes such as digital libraries, information retrieval, database systems, text/web mining, data analytics, data science, library automation, metadata, NLP, smart libraries, AI tools, research data management, and data governance.
| component | definition | use |
|---|---|---|
| Explicit AI/ML integration | AI, machine learning, neural networks, AI tools, smart libraries, NLP, text mining, web mining, generative AI, algorithmic evaluation, or AI governance is named directly. | Used to identify direct AI curriculum presence and high-integration cases. |
| Data/digital/analytics pathway | Data analytics, data science, data mining, digital libraries, research data management, social media analytics, library analytics, or information visualization is visible. | Used to capture substantial AI-adjacent preparation without equating it with explicit AI instruction. |
| AI-adjacent foundations | ICT, programming, databases, metadata, information retrieval, library automation, systems analysis, web technologies, or statistics is visible. | Used to identify enabling foundations that may support later AI learning. |
| Competency emphasis | The dominant professional capability implied by the evidence, such as technical fluency, data practice, service design, ethical evaluation, governance, or information literacy instruction. | Used to answer which competencies are emphasized across curricula. |
| Curriculum gap | A missing or weakly visible area when compared with AI literacy, responsible AI, and LIS professional competency frameworks. | Used to identify gaps such as AI ethics, generative AI, bias, privacy, AI-assisted reference, and governance. |
Table 2 summarizes the operational coding scheme. The table is intentionally compact because the study’s purpose is exploratory public-document mapping rather than exhaustive syllabus evaluation. The three integration categories support RQ2, while the competency emphasis and curriculum gap fields support RQ3 and RQ4. The distinction between explicit AI/ML integration, data/digital/analytics pathways, and AI-adjacent foundations also prevents overcounting basic ICT content as AI while still recognizing that AI preparation depends on prior digital and data competencies.
Evidence confidence was recorded separately from integration intensity. A high-confidence signal comes from a syllabus, catalogue, prospectus, curriculum plan, or course description with enough detail to support coding. A moderate-confidence signal comes from an official program page, specialization label, or course list. A low-confidence signal comes from broad competency, career, or promotional language that suggests a relationship to AI-related preparation but does not provide course-level detail. Confidence scores do not measure curriculum quality; they measure how much public evidence supports a coding decision.
3.6 Scoring and thematic coding
Each program record received an AI integration score from 1 to 3. A score of 1 indicates AI-adjacent foundations, such as ICT, information retrieval, databases, programming, systems analysis, statistics, library automation, or digitization. A score of 2 indicates a stronger data, digital, or analytics pathway, such as data science, data analytics, data mining, digital libraries, research data management, library analytics, digital humanities, or information visualization. A score of 3 indicates explicit AI/ML integration, including AI-named courses, machine learning, neural networks, NLP, text/web mining, smart libraries, or AI specialization.
Thematic coding was then applied to identify specific topic families and competency emphases. Topic codes were not mutually exclusive: a single course could be coded for both digital libraries and data analytics, or for both text mining and machine learning. Competency codes were interpretive and were assigned based on the dominant capability implied by the evidence. For example, “Text and Web Mining” was coded as text mining, web mining, NLP, and machine learning; “Digital Libraries and Resources” was coded as digital libraries and digital resource organization; and “Artificial Intelligence (AI) Application in Libraries” was coded as explicit AI in libraries.
3.7 Analytical procedures
Analysis proceeded in five steps. First, the program-level dataset was summarized by country, public evidence confidence, and AI integration score to describe the harvested evidence base. Second, course-level and course-like signals were summarized by AI relevance level and topic code to answer what AI-related topics appear in the harvested curricula. Third, competency emphases were interpreted by reading course/program evidence against AI literacy and professional competency frameworks (Association of College and Research Libraries, 2025; International Federation of Library Associations and Institutions, 2025; Lo, 2025). Fourth, curriculum gaps were identified by comparing visible curriculum signals with the professional requirements implied by the literature: technical understanding, ethical evaluation, bias and privacy awareness, data governance, generative AI use, AI-supported information literacy instruction, and AI-enabled service design. Fifth, a sensitivity check compared findings under strict, moderate, and broad definitions of AI-related curriculum content.
The sensitivity check used three assumptions. The strict definition counted only explicit AI, machine learning, NLP, text/web mining, generative AI, neural networks, smart libraries, or AI governance. The moderate definition added data mining, data analytics, data science, information extraction, sentiment analysis, information visualization, digital libraries, and research data management. The broad infrastructure definition also included ICT, databases, metadata, programming, information retrieval, systems analysis, web technologies, and automation. Descriptive counts are used only as indicators of the harvested sample. They are not population estimates for all ASEAN LIS programs. The study therefore reports frequencies as patterns within the seed dataset and pairs them with qualitative interpretation of the documentary evidence.
3.8 Trustworthiness and reproducibility
Several procedures were used to strengthen trustworthiness. First, the dataset separates source status, source URL, access date/source year, verification priority, and evidence confidence so that each coded claim can be traced back to public evidence. Second, the analysis distinguishes primary institutional documents from secondary studies. Third, coding categories are defined explicitly in Table 2. Fourth, ambiguous cases were memoed during harvesting so that digital, data, and AI-specific signals were not collapsed into a single undifferentiated “AI-related” category. Fifth, the Quarto manuscript reads directly from the harvested CSV files and generates tables and figures with R, supporting reproducibility of the descriptive summaries.
Because this is a single-researcher seed mapping exercise, formal intercoder reliability has not yet been calculated. This is a limitation. If the study is expanded into a full submission dataset, a second coder should independently code a subset of records, reconcile disagreements, and report an appropriate reliability measure. Krippendorff’s reliability guidance is relevant for that later stage because it emphasizes the need to evaluate agreement in content analysis rather than assuming that coding consistency is self-evident (Krippendorff, 2004).
3.9 Ethical considerations
The study uses publicly available institutional, professional, and scholarly documents. It does not involve human subjects, private records, interviews, or identifiable personal data beyond publicly named institutional authors and publication metadata. Ethical care is still required because public curriculum documents can be incomplete, outdated, promotional, or differently structured across countries. For this reason, the study avoids ranking institutions as “better” or “worse” and instead interprets the dataset as evidence of visible curriculum signals. Institutions with sparse public documents are not assumed to lack AI content; they are coded as having limited public evidence.
3.10 Methodological limitations
The main limitation is public-document availability. Institutions differ in how much curriculum detail they publish, whether syllabi are accessible, whether documents are current, and whether English-language pages fully represent local-language curricula. A second limitation is comparability across program levels: the seed dataset includes undergraduate programs, graduate programs, diplomas, and specializations. A third limitation is that course titles and short descriptions cannot reveal pedagogy, assessment design, contact hours, student performance, or actual classroom practice. Finally, the scoring scale captures visible integration intensity, not instructional quality.
Despite these limitations, the method is appropriate for the study’s purpose: to establish an evidence-based baseline map of AI-related curriculum content across ASEAN LIS program records and to identify where deeper syllabus collection, translation, coding, and institutional verification are needed.
4 Results
4.1 Profile of the ASEAN curriculum dataset
The harvested dataset contains 18 ASEAN LIS program records and 25 course-level or course-like AI-related curriculum signals. The program-level dataset covers eight ASEAN countries: Brunei, Indonesia, Malaysia, Myanmar, the Philippines, Singapore, Thailand, and Vietnam. The sample is intentionally described as a harvested curriculum dataset rather than a census because public curriculum availability varies across countries and institutions. Cambodia and Laos were searched but did not yield clear public LIS curriculum records during the harvesting passes.
| Country | Program/course records | Evidence note |
|---|---|---|
| Philippines | 5 program / 12 course-signal | Multiple Philippine curriculum records, including explicit AI and data/digital pathways. |
| Malaysia | 3 program / 2 course-signal | Multiple records, but public detail varies by institution and program level. |
| Indonesia | 2 program / 2 course-signal | Multiple records, but public detail varies by institution and program level. |
| Myanmar | 2 program / 2 course-signal | Multiple records, but public detail varies by institution and program level. |
| Thailand | 2 program / 2 course-signal | Multiple records, but public detail varies by institution and program level. |
| Vietnam | 2 program / 2 course-signal | Multiple records, but public detail varies by institution and program level. |
| Brunei | 1 program / 1 course-signal | Single strong program record in the seed harvest. |
| Singapore | 1 program / 2 course-signal | Single strong program record in the seed harvest. |
Table 3 shows the uneven evidence base. The Philippines has the most program records in the seed dataset, partly because several public BLIS prospectuses and program pages were discoverable. Malaysia, Thailand, Vietnam, Indonesia, and Myanmar are represented by multiple records, but the level of detail varies. Singapore and Brunei are represented by single records, yet those records are analytically important because they show strong AI/data signals: NTU Singapore offers explicit analytics and mining content, while Politeknik Brunei foregrounds library informatics, digital libraries, and data analytics.
4.2 Evidence confidence and public-document comparability
Public-document availability is not merely a limitation of the study; it is one of the findings. The same coding decision carries different interpretive weight depending on whether it is supported by a curriculum catalogue, a course description, a prospectus, a program page, or a broad specialization label. To avoid treating thin public evidence as equivalent to detailed curriculum evidence, each program record was assigned an evidence-confidence score.
| Evidence confidence | Program records | Interpretation |
|---|---|---|
| 3 - High | 6 | Coding is supported by curriculum-level public evidence such as catalogue, prospectus, or curriculum plan. |
| 2 - Moderate | 12 | Coding is supported by an official program page or comparable public program evidence. |
Table 4 shows why this study avoids ranking countries or institutions. A high integration score based on a short public page is not directly comparable with a lower integration score based on a detailed catalogue, and the absence of a visible topic in a public page cannot be read as absence from enacted teaching. The evidence-confidence variable therefore moderates all curriculum-gap claims.
4.3 Extent of AI integration across ASEAN LIS programs
AI integration was scored on a three-point scale: 1 for AI-adjacent foundations, 2 for data/digital/analytics pathways, and 3 for explicit AI/ML integration. Across the 18 program records, the most common pattern is not explicit AI instruction but visible data, digital, and analytics preparation.
Figure 3 shows that explicit AI/ML integration appears in three program records: the University of Southern Mindanao BLIS curriculum, NTU Singapore’s MSc Information Studies, and VNU-USSH Hanoi’s Digital Library and Artificial Intelligence specialization. More common are score-2 pathways where AI-adjacent preparation is carried by data analytics, digital libraries, data science, digital humanities, research data management, library analytics, and information visualization. Score-1 records remain important because they provide enabling foundations in information retrieval, databases, programming, systems analysis, library automation, and ICT, but they do not make AI itself visible.
4.5 Competencies emphasized
The harvested records emphasize technical and data-oriented competencies more strongly than critical, ethical, or governance-oriented competencies. Visible competencies include information retrieval, database design, data analytics, data mining, digital library development, web technologies, digital resource organization, and systems analysis. Explicit service and governance language is less common, although VNU-USSH’s AI specialization references AI/data governance and knowledge management, and professional frameworks indicate that ethical evaluation and responsible use are central to AI literacy (Association of College and Research Libraries, 2025; International Federation of Library Associations and Institutions, 2025).
| Competency area | Visibility | Evidence pattern |
|---|---|---|
| Technical and systems fluency | High | Information retrieval, databases, programming, web technologies, systems analysis, automation, and ICT appear across multiple countries. |
| Digital library and digital resource work | High | Digital libraries, digitization, digital repositories, digital publishing, and digital resource management recur in undergraduate, graduate, and diploma records. |
| Data and analytics practice | Moderate to high | Data analytics, data science, data mining, library analytics, social media analytics, bibliometrics, and information visualization are visible but unevenly distributed. |
| Explicit AI/ML capability | Selective | AI, machine learning, neural networks, NLP, text/web mining, and smart library services appear in a small number of high-signal records. |
| Ethics, governance, and critical AI literacy | Low to moderate | Governance and ethics appear mainly through professional frameworks or selected AI/data governance language, not consistently as named curriculum content. |
Table 6 shows that the dominant visible curriculum emphasis is operational and technical. This is not a weakness by itself: AI-related practice depends on strong foundations in retrieval, data, metadata, systems, and digital collections. However, the imbalance becomes important when compared with AI literacy frameworks that emphasize ethical, critical, and societal dimensions (Ali & Richardson, 2025; Lo, 2025). The results therefore suggest that the selected public records make the infrastructure side of AI readiness more visible than the governance and critical-literacy side.
4.6 Curriculum gaps
The gap analysis compares visible curriculum signals with the professional requirements identified in the literature and competency frameworks. The most consistent gaps are not basic digital skills but explicit AI literacy, generative AI, algorithmic bias, privacy, intellectual property, AI governance, and AI-supported service design.
| Gap area | Visibility | Curriculum implication |
|---|---|---|
| Generative AI in LIS practice | Weakly visible | Curricula should specify how generative AI affects reference, information literacy instruction, metadata, research support, and scholarly communication. |
| Algorithmic bias and evaluation | Weakly visible | Programs need explicit outcomes for evaluating AI outputs, bias, transparency, hallucination, and reliability. |
| Privacy, intellectual property, and data governance | Selective | Responsible AI should be connected to privacy, copyright, data stewardship, and institutional governance. |
| AI-enabled service design | Selective | Smart library services and user analytics should be paired with human-centered service design and assessment. |
| Practical AI/ML/NLP labs using LIS data | Selective | Technical courses could use library metadata, retrieval logs, digital collections, bibliographic data, or archival corpora for applied learning. |
Table 7 identifies the gap pattern most relevant for curriculum revision. The main issue is not simply that AI content is absent. Rather, AI content appears unevenly and is often mediated through adjacent technical or digital topics. The gap is therefore one of explicit alignment: programs need to clarify whether AI-related courses are intended to develop tool use, data competence, computational understanding, critical AI literacy, responsible governance, or AI-informed service design.
4.7 Sensitivity analysis: strict, moderate, and broad AI definitions
Because “AI-related content” can be defined too narrowly or too broadly, the study tested whether the interpretation changes under three coding assumptions. This sensitivity check is especially important for Option A because the dataset is small and public-document visibility is uneven.
| Coding assumption | Course signals | Interpretation |
|---|---|---|
| Strict AI definition | 6 | Counts only explicit AI/ML/NLP/text-web mining/generative AI/governance signals. |
| Moderate AI-adjacent definition | 15 | Adds data mining, analytics, digital libraries, RDM, visualization, and data governance. |
| Broad infrastructure definition | 23 | Adds broader ICT, retrieval, database, metadata, programming, systems, and automation foundations. |
Table 8 shows that the study’s conclusions depend on definitional breadth. Under a strict definition, explicit AI/ML signals are selective. Under a moderate definition, data, analytics, text mining, digital libraries, and governance-related pathways make AI preparation more visible. Under a broad infrastructure definition, many more LIS technology foundations become relevant, but they should not be interpreted as direct AI instruction. The most defensible conclusion is therefore not that ASEAN LIS programs are uniformly AI-ready or AI-lacking. It is that public curriculum evidence shows a spectrum from enabling infrastructure to explicit AI/ML integration, with responsible AI and critical AI literacy less consistently visible.
4.8 Summary of findings by research question
Taken together, the results answer the five research questions as follows. For RQ1, the AI-related topics visible in the harvested public evidence include artificial intelligence, machine learning, NLP, text/web mining, data mining, data analytics, data science, digital libraries, information retrieval, database systems, web technologies, systems analysis, digital humanities, research data management, and library analytics. For RQ2, integration is uneven: score-2 data/digital/analytics pathways are more common than explicit score-3 AI/ML integration. For RQ3, visible curriculum signals emphasize technical, systems, digital library, and data competencies more strongly than ethics, governance, and critical AI literacy. For RQ4, the most visible alignment gaps concern generative AI, algorithmic bias, privacy and intellectual property, AI governance, human-centered AI service design, and applied AI labs using LIS datasets. For RQ5, public-document availability shapes comparability: stronger curriculum documents support more confident coding, while thin program pages require more cautious interpretation.
These findings should be interpreted as a baseline map rather than a final judgment on ASEAN LIS education. Public evidence shows that AI-relevant preparation is already present, but often under different labels. The strategic curriculum task is to make those connections explicit, sequence them coherently, and align them with professional expectations for responsible AI practice in libraries and information organizations.
5 Discussion
5.1 Reading visible AI integration as a curriculum spectrum
The results suggest that visible AI integration in the selected ASEAN LIS records is better understood as a spectrum than as a simple presence-or-absence condition. Explicit AI/ML integration appears in a small number of program records, but AI-adjacent preparation is more widely distributed through digital libraries, information retrieval, database systems, data analytics, data mining, data science, systems analysis, and digital resource management. This finding is important because contemporary AI work in libraries depends on infrastructures and competencies that are not always named as AI. Retrieval systems, metadata, digital collections, analytics, and data governance all shape whether library workers can understand, evaluate, and responsibly use AI-enabled systems.
This spectrum interpretation is consistent with the reviewed literature. Cox (2021) and Cox (2023) frame AI as a professional transformation that affects information work through systems, services, expertise, and organizational judgment. The ASEAN library-adoption literature similarly shows that AI uptake is uneven across countries and institutions, shaped by policy support, infrastructure, and organizational readiness (C. Xu & Loo, 2025; Y.-S. Xu et al., 2025). The public curriculum evidence therefore mirrors the professional landscape: AI is emerging, but it is not yet consistently formalized as a named curriculum object across the selected ASEAN LIS program records.
5.2 Curriculum Theory interpretation
From the perspective of Curriculum Theory, the findings point to a gap between visible curriculum content and the emerging professional meaning of AI in librarianship. The formal curriculum already contains important foundations, especially in systems, databases, retrieval, digital libraries, and data-oriented work. However, the formal curriculum often leaves the relationship between these foundations and AI professional practice implicit. This matters because curriculum is not only a list of courses; it is a structured argument about what knowledge matters, how knowledge should be sequenced, and what kinds of professional identity a program is trying to form.
The visible curriculum pattern suggests that many ASEAN LIS programs are still organizing AI-relevant learning through older curricular categories: ICT, automation, digital libraries, information retrieval, and data management. These categories remain necessary, but they do not fully communicate the newer professional questions raised by AI: how algorithmic systems shape access, how generated content affects authority and trust, how data-intensive systems interact with privacy and intellectual property, and how librarians should teach users to evaluate AI tools. In Curriculum Theory terms, the issue is not simply content addition. It is curriculum rearticulation: programs need to make explicit how established LIS foundations now connect to AI-mediated information environments.
5.3 Competency-Based Education implications
The competency pattern identified in the Results section shows strong emphasis on technical and operational readiness, but weaker visibility for ethical, critical, governance, and instructional competencies. CBE makes this imbalance consequential because it asks whether curriculum evidence aligns with the capabilities graduates are expected to demonstrate in practice. Recent professional frameworks increasingly define AI readiness in multidimensional terms. ACRL’s AI competencies emphasize knowledge, evaluation, ethical awareness, and applied use, while IFLA frames AI in relation to rights, equity, literacy, privacy, and professional responsibility (Association of College and Research Libraries, 2025; International Federation of Library Associations and Institutions, 2025). Lo (2025) similarly treats AI literacy as technical, ethical, critical, and societal rather than merely instrumental.
The ASEAN curriculum evidence therefore suggests a competency alignment problem. Students may receive useful preparation for the technical substrate of AI, but the public records do not consistently show explicit learning outcomes for evaluating AI outputs, detecting bias, explaining hallucination and uncertainty, protecting user data, assessing vendor claims, designing AI-supported services, or teaching AI literacy. This matters because library workers are increasingly positioned not only as users of AI tools, but also as mediators who help students, researchers, faculty, and communities understand AI systems. Studies of LIS students and academic-library LibGuides reinforce this point: AI literacy is being treated as both a curriculum concern and an instructional service concern, with persistent need for stronger critical, ethical, and technical scaffolding (Hossain et al., 2025; Ko & Tang, 2025). As Bridges et al. (2025) and Ali & Richardson (2025) argue, AI literacy in LIS education and academic library practice requires critical and policy-aware competencies, not only tool familiarity.
5.4 ASEAN regional significance
The unevenness found in the curriculum map should be interpreted in relation to ASEAN’s wider policy and institutional context. The ASEAN Responsible AI Roadmap 2025–2030 emphasizes governance, capacity building, infrastructure, and regional cooperation (ASEAN Secretariat, 2025). At the same time, ASEAN LIS education has long faced questions of quality assurance, regionalization, and uneven institutional capacity (Sacchanand, 2015). The present results sit at the intersection of these two agendas. If responsible AI is becoming a regional capacity-building priority, LIS programs have a plausible role in preparing professionals who can organize information, support research, teach information literacy, steward data, and evaluate algorithmic systems.
The findings also caution against a one-size-fits-all curriculum recommendation. Singapore, Malaysia, the Philippines, Vietnam, Thailand, Indonesia, Brunei, and Myanmar appear in the seed dataset with different levels of public curriculum detail and different kinds of AI-relevant evidence. Some programs show explicit AI, text mining, machine learning, or AI specialization signals. Others show foundational digital and systems preparation. A regional benchmarking tool should therefore recognize multiple entry points. For programs with strong ICT foundations, the next step may be explicit AI literacy and governance outcomes. For programs with data or analytics pathways, the next step may be applied LIS datasets, human-centered service design, and critical evaluation. For programs with explicit AI courses, the next step may be integration across the curriculum rather than isolating AI in a single elective.
5.5 Explaining the curriculum gaps
The gaps identified in Table 7 are best understood as alignment gaps rather than total absences. Generative AI, algorithmic bias, privacy, intellectual property, AI governance, AI-supported service design, and applied AI labs are weakly or selectively visible in the harvested public evidence. Yet these areas are central to contemporary library AI practice because libraries are already encountering AI in discovery, scholarly communication, reference, research support, metadata workflows, teaching, and vendor platforms (Cox, 2023; International Federation of Library Associations and Institutions, 2025). The lack of explicit public curriculum language may make it harder for programs to demonstrate that graduates can handle these responsibilities.
One explanation is that AI has moved faster than formal curriculum revision cycles. Public curricula often lag behind practice because degree revisions require institutional approval, faculty capacity, regulatory alignment, and sometimes professional accreditation review. Recent LIS education work explicitly frames emerging topics and new course development as matters of strategic curriculum planning rather than simple topic insertion (Chen et al., 2024). Another explanation is that AI-related learning may be occurring informally through special topics, workshops, faculty initiatives, practicum experiences, or course-level assignments that are not visible in public documents. This possibility reinforces a methodological caution: public curriculum mapping can identify visible content and visible silence, but it cannot fully capture enacted teaching without syllabi, assignments, classroom observation, or faculty interviews.
5.6 Toward strategic curriculum alignment
The practical implication is that ASEAN LIS schools do not necessarily need to replace existing digital, systems, and data courses. Instead, they can strengthen AI readiness by adding explicit alignment across existing curriculum structures. Information retrieval can include algorithmic ranking, evaluation, explainability, and bias. Metadata and cataloging can include AI-assisted description, provenance, quality control, and labor implications. Digital libraries can include machine learning for discovery, user analytics, preservation risks, and rights management. Reference and information literacy can include generative AI evaluation, prompt practices, hallucination, citation integrity, and user education. Management and policy courses can include AI procurement, privacy, governance, accountability, and institutional guidelines.
This approach fits both Curriculum Theory and CBE. Curriculum Theory supports the reorganization of existing knowledge into a coherent professional formation, while CBE requires explicit outcomes that can be assessed. A strategic ASEAN LIS AI curriculum framework would therefore include sequenced competencies: foundational digital and data fluency, applied AI/ML awareness using LIS-relevant examples, critical AI literacy, responsible governance, and human-centered AI service design. This sequencing also aligns with curriculum modernization arguments that connect LIS education to industrial and technological transformation (Rahmah & Marlini, 2020). Such a framework would help programs avoid two weak options: treating AI as only a technical elective, or treating it as a vague emerging issue without assessable outcomes.
5.7 Implications for research and benchmarking
For research, the study demonstrates that a harvested public-document dataset can support a first-stage regional curriculum map. The next stage should deepen the evidence base by adding full syllabi, learning outcomes, assessment tasks, internship or practicum requirements, and faculty-verified course descriptions. Reliability procedures can also be strengthened through multiple coders, intercoder agreement testing, and iterative refinement of the coding scheme (Krippendorff, 2004, 2019). These steps would move the study from an exploratory seed map toward a stronger benchmarking instrument.
For benchmarking, the most useful comparison is not a ranking of countries or programs. A better benchmark is a competency matrix that shows where AI-related preparation appears, how explicit it is, what competencies it supports, and what gaps remain. Such a tool would allow LIS schools to compare their curriculum against regional peers and against professional expectations without assuming that every institution must adopt the same course titles. It would also support incremental curriculum development, allowing programs to identify whether they need an introductory AI literacy module, an applied AI elective, stronger data governance content, or AI-infused revisions across existing courses.
6 Recommendations and Conclusion
6.1 Tiered curriculum recommendations
The findings support tiered recommendations rather than a single regional prescription. Programs whose public records mainly show ICT, retrieval, databases, metadata, programming, automation, or systems foundations can strengthen AI readiness by adding explicit AI literacy, algorithmic evaluation, privacy, bias, and responsible-use outcomes to existing courses. Programs with visible data, digital library, analytics, data mining, digital humanities, or research data management pathways can add LIS-specific applied AI activities using metadata, retrieval logs, digital collections, bibliographic datasets, reference transcripts, or archival corpora. Programs with explicit AI, machine learning, NLP, text mining, or AI specialization signals can move beyond isolated electives by integrating responsible AI, user education, governance, and service-design questions across reference, metadata, digital libraries, management, and information literacy instruction. Programs with sparse public documentation should prioritize curriculum transparency before being included in stronger regional benchmarking.
Across all program types, curriculum revision should make competency alignment explicit. A course title such as “Digital Libraries” or “Information Retrieval” may provide strong AI-relevant foundations, but it should state whether students are expected to evaluate ranking systems, understand machine learning applications, manage data ethics, assess vendor tools, or teach users about generative AI. The practical curriculum goal is not to relabel every digital course as AI. It is to clarify where AI-specific, AI-adjacent, and responsible AI competencies are introduced, practiced, assessed, and reinforced.
6.2 Future research
Future research should extend this pilot in four ways. First, a verified ASEAN LIS population frame should be constructed using national higher education directories, professional association lists, university catalogues, and multilingual search procedures. Second, program records should be supplemented with syllabi, learning outcomes, assignments, practicum requirements, and faculty-verified course descriptions. Third, reliability should be strengthened through at least two coders, a formal codebook, independent coding of a subset of records, disagreement reconciliation, and a reported intercoder agreement statistic where appropriate. Fourth, sensitivity analysis should be retained so that readers can see how conclusions change under strict, moderate, and broad definitions of AI-related curriculum.
6.3 Conclusion
This article has reconstructed the study as an exploratory public-document mapping of AI-related curriculum visibility in selected ASEAN LIS program records. The evidence shows a curriculum spectrum: explicit AI/ML integration is visible in a small number of records, while data, digital library, analytics, information retrieval, database, programming, and systems foundations appear more widely. The most important curriculum issue is not simply whether AI is present or absent. It is whether public curriculum evidence makes AI-related competencies explicit enough to support responsible professional formation.
The study’s strongest contribution is therefore methodological and developmental. It offers a bounded way to map visible curriculum signals, classify integration intensity, compare signals with professional competency frameworks, and report uncertainty created by uneven public documentation. Used cautiously, this framework can help LIS programs identify where existing digital and data courses already support AI readiness, where responsible AI and critical AI literacy need stronger visibility, and where future research must move beyond public documents to verified syllabi, faculty perspectives, and evidence of student learning.
References
Ali, M. Y., & Richardson, J. (2025). AI literacy guidelines and policies for academic libraries: A scoping review. IFLA Journal, 51(3), 588–599. https://doi.org/10.1177/03400352251321192
Arafeh, S. (2016). Curriculum mapping in higher education: A case study and proposed content scope and sequence mapping tool. Journal of Further and Higher Education, 40(5), 585–611. https://doi.org/10.1080/0309877X.2014.1000278
ASEAN Secretariat. (2025). ASEAN responsible AI roadmap 2025–2030. https://asean.org/wp-content/uploads/2025/02/ASEAN-Responsible-AI-Roadmap-Final.docx.pdf
Association of College and Research Libraries. (2025). AI competencies for academic library workers. https://www.ala.org/acrl/standards/ai
Bowen, G. A. (2009). Document analysis as a qualitative research method. Qualitative Research Journal, 9(2), 27–40. https://doi.org/10.3316/QRJ0902027
Bridges, L. M., Lopezosa, C., Centelles Velilla, M., & Ferran-Ferrer, N. (2025). Moving towards critical AI literacy in LIS education: A scoping review and syllabi analysis. The Electronic Library, 43(5), 798–817. https://doi.org/10.1108/EL-06-2025-0224
Chen, J., Taylor, N., & Rorissa, A. (2024). Emerging topics, challenges, and strategies for library and information science education. Proceedings of the ALISE Annual Conference. https://doi.org/10.21900/j.alise.2024.1693
Chulalongkorn University Department of Library Science. (2026). Bachelor of arts program in information studies. https://www.arts.chula.ac.th/libsci/en/curriculum/undergraduate/
Cox, A. M. (2021). The impact of AI, machine learning, automation and robotics on the information professions: A report for CILIP. CILIP: The Library; Information Association. https://eprints.whiterose.ac.uk/id/eprint/174522/
Cox, A. M. (2023). How artificial intelligence might change academic library work: Applying the competencies literature and the theory of the professions. Journal of the Association for Information Science and Technology, 74(3), 367–380. https://doi.org/10.1002/asi.24635
Hossain, Z., Biswas, Md. S., & Khan, G. (2025). AI literacy of library and information science students: A study of bangladesh, india and pakistan. Journal of Librarianship and Information Science, 58(1), 285–299. https://doi.org/10.1177/09610006241309323
International Federation of Library Associations and Institutions. (2025). IFLA statement on libraries and artificial intelligence. https://repository.ifla.org/items/8c05d706-498b-42c2-a93a-3d47f69f7646
Ko, C. R., & Tang, R. (2025). Promoting AI literacy through u.s. Academic libraries: An analysis of LibGuides from ARL and oberlin group libraries using the EDUCAUSE AI literacy framework. Information Research: An International Electronic Journal, 30(iConf), 847–865. https://doi.org/10.47989/ir30iConf47182
Krippendorff, K. (2004). Reliability in content analysis: Some common misconceptions and recommendations. Human Communication Research, 30(3), 411–433. https://doi.org/10.1111/j.1468-2958.2004.tb00738.x
Krippendorff, K. (2019). Content analysis: An introduction to its methodology (4th ed.). SAGE Publications. https://methods.sagepub.com/book/mono/content-analysis-4e/toc
Lo, L. S. (2025). AI literacy: A guide for academic libraries. College & Research Libraries News, 86(5). https://crln.acrl.org/index.php/crlnews/article/view/26704/34626
Nanyang Technological University. (2026). Master of science in information studies. https://www.ntu.edu.sg/education/graduate-programme/master-of-science-in-information-studies
Politeknik Brunei. (2026). Diploma in library informatics computing. https://pb.edu.bn/politeknik-brunei/school-of-information-and-communication-technology/diploma-in-library-informatics-computing/
Rahmah, E., & Marlini. (2020). Curriculum development of library and information science study program in 4.0 industrial revolution era. Proceedings of the 4th International Conference on Language, Literature, Culture, and Education. https://doi.org/10.2991/assehr.k.201109.059
Sacchanand, C. (2015). Quality assurance of library and information science education in the ASEAN countries: Moving towards regionalization and internationalization. World Library and Information Congress: 81st IFLA General Conference and Assembly. https://repository.ifla.org/items/1f38c568-e5f6-47f8-b6b5-58d44f0c8b10/full
Subaveerapandiyan, A., Paladhi, M. M., & Maruthaveeran, V. (2024). Evaluating AI literacy proficiency among LIS researchers in ASEAN. Library Hi Tech News, 41(4), 6–8. https://doi.org/10.1108/LHTN-07-2023-0121
Universitas Indonesia Faculty of Humanities. (2025). Buku pedoman sarjana 2025/2026. https://fib.ui.ac.id/wp-content/uploads/sites/8/2025/08/BUKU-PEDOMAN-SARJANA-2025.pdf
Universiti Teknologi MARA. (2026). SI260 bachelor of information science (honours) library management. https://fis.uitm.edu.my/academics/undergraduate-degree/si260
University of Social Sciences and Humanities, Vietnam National University Hanoi. (2025). Digital library and artificial intelligence specialization. https://ussh.vnu.edu.vn/eng/dao-tao/nganh-thong-tin-hoc/chuyen-nganh-thu-vien-so-tri-tue-nhan-tao-bat-nhip-cung-ky-nguyen-so-23080.html
University of Southern Mindanao. (2024). Bachelor of library and information science curriculum. https://www.usm.edu.ph/academics/colleges/college-of-engineering-and-information-technology/
Xu, C., & Loo, S. (2025). A review of artificial intelligence applications in libraries in southeast asia: Where are we now? Reference Services Review, 53(1), 66–91. https://doi.org/10.1108/RSR-06-2024-0027
Xu, Y.-S., Yanti Idaya, A. M. K., & Kassim, M. S. S. (2025). From vision to practice: Adoption of artificial intelligence in libraries across ASEAN. Malaysian Journal of Library & Information Science, 30(2). https://doi.org/10.22452/mjlis.vol30no2.3