Saudi AI research mapped against Vision 2030 reveals striking gaps

A new bibliometric review of artificial intelligence scholarship in Saudi Arabia has found that academic output in the kingdom, while growing rapidly, covers only part of the innovation agenda that Vision 2030 sets out. The study, published in Discover Artificial Intelligence, analyzed 75 peer-reviewed articles published between 2015 and 2025 and retrieved from the Web of Science database, combining bibliometric mapping with a systematic literature review. Its central finding is a measurable alignment gap: several sectors that the national strategy explicitly targets, including tourism, mining, retail, logistics, and pilgrimage management, produced no distinct research clusters at all.

The research team, led by Zeineb Al Hadi Nagara with Ahmad Haidar and Henry Hassan Ayoub, applied a ten-step bibliometric-systematic literature review methodology. The authors frame their work as the first cross-sectoral, strategy-aligned mapping of Saudi AI research, designed to bridge what they call the strategic alignment gap between what scholars study and what national innovation policy actually needs. The keyword co-occurrence analysis, using author keywords with a minimum threshold of two occurrences, identified eight thematic clusters. To assess alignment, the team also developed a three-category coding scheme, distinguishing direct programmatic alignment, sectoral alignment, and strategic gaps, with two researchers coding independently and reconciling through discussion.

The eight clusters are core AI and economic readiness; AI in healthcare and crisis response; information management for data-driven decision-making; innovation and sustainability; national strategy and policy orientation; bibliometric and higher education research; AI in public health and decision support; and a small, isolated cluster on the banking sector. The keyword artificial intelligence occurred 29 times in the network and showed strong conceptual links to machine learning and to Saudi Arabia itself, while the banking sector keyword, though placed in its own cluster, co-occurred strongly with those same terms. The authors report that the cluster structure proved robust: re-running the analysis at higher thresholds of three and four occurrences preserved the same intellectual anchors, with eight keywords surviving at threshold three and six at threshold four.

Publications on AI and economic transformation in Saudi Arabia first appeared only in 2020, five years into the study window, reflecting a lag between policy announcement and scholarly response. Early work between 2020 and 2022 concentrated on foundational themes such as Industry 4.0, user acceptance, governance, and pandemic response. A middle period from 2022 to 2023 shifted toward machine learning, big data, adoption, and economic growth, while research after 2023 turned to specific applications including diagnosis, digital economy, optimization, sustainability, and water resources. The authors interpret this trajectory as evidence that research priorities have tracked the evolving Vision 2030 agenda, a pattern they link to Dynamic Capabilities Theory, in which organizations sense policy shifts and reconfigure around new demands.

The healthcare clusters illustrate how policy context has shaped research. Vision 2030’s Health Sector Transformation Program recognized AI as a key enabler for addressing the kingdom’s rising rates of non-communicable disease: 18 percent of the adult population is currently afflicted by diabetes, and the elderly population is growing at 3.2 percent per annum, projected to reach 25 percent of the total population by 2050. The COVID-19 pandemic acted as a further catalyst, with studies documenting AI’s role in the national response, including deep learning models that improved the accuracy of case forecasting and AI-supported surveillance and resource coordination. The crisis-response cluster also extends to environmental monitoring, with AI-based models used to map groundwater vulnerability and to predict changes in terrestrial water storage from GRACE satellite data, reflecting the strategic weight of water scarcity in an arid economy.

Institutional analysis through bibliographic coupling showed a highly uneven distribution of research activity. Four institutions exceeded a total link strength of 500 and formed the high-engagement tier: Imam Abdulrahman Bin Faisal University, Prince Sultan University, Lahore University of Management Sciences, and University of the Punjab. Notably, two of the four most central institutions are Pakistani rather than Saudi. A moderate tier, with total link strength between 200 and 500, included Princess Nourah Bint Abdulrahman University, University of Hail, Saudi Electronic University, King Fahd University of Petroleum and Minerals, and King Saud University, while institutions such as Qatar University and Taibah University showed low network integration below 200.

One of the more surprising findings concerns regional collaboration. Saudi Arabia’s strongest bibliographic ties were with geographically distant countries: Australia with a total link strength of 524, India at 511, Pakistan at 495, and Japan at 329. By contrast, its links with neighboring Gulf states were far weaker, with the United Arab Emirates at 103 and Kuwait at 40. The authors suggest this paradox may reflect nationally isolated research goals, linguistic barriers, or a preference for Arabic-language or local publications not indexed in Web of Science. They argue the result challenges the assumption, embedded in National Innovation Systems thinking, that geographic proximity naturally fosters knowledge flows in emerging-economy innovation systems, suggesting instead that policy alignment and disciplinary orientation are stronger predictors of knowledge connectivity in state-led contexts.

The systematic review component identified five cross-cutting themes across the corpus: responsible and ethical AI development, corporate digital responsibility, AI’s contribution to the Sustainable Development Goals, AI regulation, and workforce upskilling. Studies linked AI applications to SDG 3 on good health and well-being, SDG 7 on affordable and clean water, SDG 8 on decent work and economic growth, and SDG 11 on sustainable cities and communities. The review also found that ethical principles such as transparency and explainability remain more of an aspiration than a reality at the organizational level, and that regulatory ambiguities persist around system interoperability, data privacy, and responsible use in sensitive domains such as medical education. On upskilling, the workforce literature was concentrated exclusively in the health and education sectors, leaving most business functions without a scholarly roadmap for AI-driven workforce development.

The strategic gaps the study documents are extensive. Beyond the absence of research clusters on tourism, mining, retail, logistics, and pilgrimage management, the authors highlight several underdeveloped technology domains. Generative AI and large language models are described as the most conspicuous absence, despite their potential for Arabic language processing, government service automation, and educational content generation. Digital twin technology has appeared only in construction research, and explainable AI, which bears directly on the kingdom’s stated ambition for transparent and accountable AI governance, has received minimal attention. The authors also note unexplored intersections between AI and Islamic finance, and between AI and cultural heritage digitization aligned with Vision 2030’s quality-of-life objectives.

The study documents the scale of the national AI ambition against which this research output should be read. Saudi Arabia’s National Strategy for Data and AI, launched in 2020, projects 135 billion dollars in AI-related revenue by 2030 and targets the training of 2,000 AI specialists. The kingdom scored 72.36 on the Oxford Insights Government AI Readiness Index in 2024, second in the region behind the United Arab Emirates at 75.66, and topped Tortoise Media’s Government AI Strategy Index the same year. Sixty-six of the 96 Vision 2030 objectives are directly or indirectly related to data and AI. The partnership between the Saudi firm Humaine and NVIDIA, involving 18,000 advanced AI chips for data centers with a combined capacity of 500 megawatts, illustrates the scale of infrastructure investment, and a Cloud Computing Special Economic Zone established in April 2023 is anticipated to capture 30 percent of total ICT spend by 2030 and add 7 billion Saudi riyals to GDP.

Quantitative outcomes elsewhere in the Vision 2030 agenda also frame the mapping. Electronic payments accounted for more than 57 percent of retail transactions by the end of 2021, beating the initial cashless-society target ahead of schedule, and homeownership rose from 47 percent to 65.4 percent by the end of 2024, increasing demand for AI-based targeting, service automation, and fraud detection. Yet the banking research cluster contains only two studies, which the authors identify as a research void in a strategically significant sector under the Financial Sector Development Program. Citation analysis reinforced the field’s orientation: the most cited article, by Baabdullah and colleagues on AI-based B2B practices in Saudi SMEs, had 95 citations, and co-citation analysis of sources revealed two clusters, one grouping high-impact management and information systems journals and the other applied and technical outlets such as IEEE Access.

The authors acknowledge several limitations. The analysis relies on a single database, Web of Science, and includes only English-language peer-reviewed journal articles, excluding conference papers, book chapters, and Arabic-language scholarship, which likely omits practitioner-oriented and regional work. VOSviewer’s modularity-based clustering depends on parameter choices, most importantly the minimum keyword occurrence threshold, and the interpretation of bibliometric clusters retains an inherently qualitative dimension that the authors say future studies could strengthen through formal inter-rater reliability testing. Despite these constraints, the mapping offers policymakers and research funders an evidence-based picture of where Saudi AI scholarship is concentrated and where it is absent, and the authors recommend formalizing a national Responsible AI Framework spanning ethical, sustainability, legal compliance, and responsible innovation management dimensions.

Subject of Research: Alignment of Saudi Arabian academic AI research with the Vision 2030 national innovation strategy

Article Title: Aligning academic artificial intelligence research with the national innovation strategy of Saudi Arabia under vision 2030

Article References: NAGARA, Z. A. H., HAIDAR, A., & AYOUB, H. H. (2026). Aligning academic artificial intelligence research with the national innovation strategy of Saudi Arabia under vision 2030. Discover Artificial Intelligence, 6(1), Article 985. https://doi.org/10.1007/s44163-026-01971-0

Image Credits: AI Generated

DOI: 10.1007/s44163-026-01971-0

Keywords: artificial intelligence, Saudi Arabia, Vision 2030, bibliometric analysis, systematic literature review, national innovation strategy, digital transformation, AI readiness, research-policy alignment, Gulf collaboration, responsible AI, workforce upskilling