INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 1369
www.rsisinternational.org
Exploring the Key Challenges Faced By the South African Public
Sector in Adopting Artificial Intelligence
Wiston Mbhazima Baloyi*
Independent Researcher, Limpopo Province, Polokwane, South Africa
DOI: https://dx.doi.org/10.47772/IJRISS.2025.910000117
Received: 06 October 2025; Accepted: 14 November 2025; Published: 05 November 2025
ABSTRACT
The widespread use of artificial intelligence (AI) technologies has not only streamlined internal processes of the
public sector but also advanced public services to citizens globally. Although these trailblazing technologies
have been valued for their enhanced efficiency and streamlined processes, they have been susceptible to
challenges that hinder their full potential, particularly in developing countries like South Africa. The purpose of
this study is to explore the key challenges faced by the South African public sector in adopting AI. Grounded in
the interpretivist paradigm, this study employs a qualitative methodology, utilising the Preferred Reporting Items
for Systematic Reviews and Meta-Analysis (PRISMA) technique to collect and synthesise data from the
literature. The systematic analysis included 20 studies extracted from the Web of Science database. The findings
of this study reveal that digital infrastructural deficit, digital illiteracy, policy and regulatory gaps, and ethical
dilemmas are all key challenges encountered in the adoption of AI in the South African public sector. The
implications of this study are relevant to deepening understanding and advising policymakers, decision-makers,
and practitioners on AI challenges.
Keywords: Artificial intelligence, Challenges, Public services, Public sector, PRISMA, South Africa
INTRODUCTION
The ubiquitous application of information and communication technology (ICT) has profoundly transformed
public services (Baloyi & Beyers, 2019). Artificial intelligence (AI) has recently gained momentum across
diverse sectors, including public administration. With the burgeoning use of AI in the global realm, public sectors
have undergone a profound reshaping of public services for citizens and various stakeholders. AI-driven
mechanisms and techniques, such as machine learning, virtual assistants, virtual payments, natural language
processing, deep learning and chatbots, among others, have tremendously transfigured the internal processes of
public sectors and improved service delivery (Tveita & Hustad, 2025). While AI is regarded as preliminary in
developing countries like South Africa (Janneker, 2025), most advanced economies across the globe (e.g., the
United Kingdom, the United States of America, Australia and Japan) have long been on the edge of automated
decision-making and already reaped the benefits brought by AI-enabled technologies (Baloyi, 2025). More to
this, whereas most advanced economies have continuously been rendering digitalised services to the citizens,
developing countries such as South Africa are still beleaguered with contextual challenges, including but not
limited to digital infrastructure, digital illiteracy, policy and regulatory deficit, and ethical dilemmas, impeding
the full-fledged usage of AI technologies (Shekgola & Modiba, 2025; Chilunjika, 2024).
South Africa has consistently been a beacon of hope in service delivery to its citizens. The Republic of South
Africa gained freedom after the apartheid regime and post-colonisation in 1994. Nevertheless, even after 31
years of democracy, the country remains plagued by socio-economic discrepancies, commonly referred to as the
triple scourge: poverty, inequality, and a high unemployment rate (Baloyi, 2025). This triple scourge is
detrimental and often impedes the complete realisation of emerging digital technologies, including AI. For
instance, although the country experiences obstacles such as a digital infrastructural deficit, unreliable internet
connectivity, and cellular network disparities, the digital divide is regarded as the most significant contributing
factor exacerbating digital failures, especially in remote or rural areas, affecting marginalised groups (Malope,
2025; Maleka & Maidi, 2024). Further to this, despite the significant strides made in embracing AI-driven
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 1370
www.rsisinternational.org
technologies, the lack of a legal and regulatory framework governing the ethical use of AI appears to be
perplexing (Naidoo, 2024). In other words, the paucity of policies regulating AI in the South African public
sector remains a challenge to the implementation of digital ideas. Even though steps were taken to embrace and
tackle AI by developing policy, such as “the South African National Artificial Intelligence Policy Framework”
in 2024, to date, however, the cabinet has not yet endorsed such a policy (Baloyi, 2025). Given these debates, it
is essential to investigate the key challenges the South African public sector faces in adopting AI to thrive in the
digital landscape.
In the South African public sector context, AI has been rampant and manifested in different forms to diverse
industries, include but not limited to, healthcare services (Janneker, 2025; Malope, 2025), municipal services
(Bester, 2024), rural development (Naidoo, 2024), public administration (Modiba, 2025), and basic education
(Saal et al., 2025), inter alia. Notwithstanding the vulnerabilities outlined earlier, AI-related technologies have
considerably increased efficiency and effectiveness in delivering services in South Africa, more especially in the
private sector. Yet, like many other countries throughout the world, South Africa is expected to achieve the
Sustainable Development Goals (SDGs) as set by the United Nations, as well as the National Development Plan
(NDP) and Vision 2030, which is fast approaching. That being the case, implementing digital technologies
(including AI) is imperative to enhance citizens’ welfare while working towards the realisation of the NDP,
aiming to eradicate poverty. Against this background, this study aims to address the following research question:
What are the key challenges faced by the South African public sector in adopting AI?
Theoretical Perspective
Several theories guide and navigate the adoption of technology at both the individual and institutional levels in
the public sector. These include, but are not limited to, technology acceptance model (TAM), technology
acceptance model (TAM) version 2, technology–organisation–environment framework (TOE), theory of
planned behavior (TPB) and unified theory of acceptance and use of technology (UTAUT) (Baker, 2012). Due
to the proliferating necessity to embrace AI to deliver streamlined public services in the digital landscape and
the novelty of its adoption in the South African public sector, Tornatzky and Fleischer's (1990) TOE is seen as
pertinent in evaluating the capabilities and limitations of institutions using the three contextual dimensions:
technology, organisation, and environment (see Figure 1 below). The TOE has been extensively employed by
diverse sectors (both private and public) in appraising organisational condition when pursuing the uptake of
digital technologies, utilising these three dimensions. Notwithstanding its wide application in diverse phenomena
where technology acceptance is prevalent, the TOE has been commended for its feasibility and practicability in
assessing the organisational capacity (e.g., the capability and availability of resources).
Figure 1 Technology-organisation-environment model (Tornatzky and Fleischer, 1990)
The technological dimension relates to the scope and rapidity of technological transformations and innovations
that are volatile, erratic, and non-negotiable, yet applicable for acceptance by the public sector (Bryan & Zuva,
2021). It includes digital advances (such as AI, cloud computing, virtual platforms, big data, and the Internet of
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 1371
www.rsisinternational.org
Things) that governments can adopt to enhance their decision-making processes and deliver efficient services to
society. On the other hand, the organisational dimension is concerned with scanning internal capacities (i.e.,
strengths and vulnerabilities) that are critical in accepting new digital technologies (Baker, 2012). It involves
having sufficient resources to accommodate technological changes, such as human, financial, and technical
resources, as well as an innovative culture, a vibrant organisational structure, leadership drive, and effective
reward strategies (Awa et al., 2017). The environmental dimension entails being mindful of the ecosystem
(ecology) while adhering to governmental prescriptions when adopting digital technologies (Baker, 2012). These
include regulatory frameworks and guidelines that direct digital technologies in the digital era.
MATERIALS AND METHODS
To achieve the purpose and respond to the research question of this study, a critical review of academic discourse
was conducted to ensure an ample, all-inclusive, explicit, and replicable method for retrieving, selecting, and
interpreting literature related to the current topic (Page et al., 2021). Grounded on the interpretivist paradigm,
this study employs qualitative methodology using the Preferred Reporting Items for Systematic Reviews and
Meta-Analysis (PRISMA) technique to explore the key challenges faced by the South African public sector in
adopting AI. The PRISMA technique has been lauded for enhancing transparency, inclusiveness, and reliability
in appraising the literature by rigorously clustering related studies within a specific area or field (Page et al.,
2021). According to Sohrabi (2021), the advantages of employing the PRISMA technique include: (1) while
executing large amounts of data, it ensures explicit and dependable study conclusions, (2) it helps reduce
prejudice in the selection and interpretation of literature, and (3) it enables critical assessment of literature and
identification of pertinent themes. Even though PRISMA originates from medical research (Page et al., 2021),
it has been extensively applied in social science research, particularly in the context of AI phenomena in the
public sector (Madan & Ashok, 2023; Tomazevic et al., 2024). The PRISMA process involves four steps:
identification, screening, eligibility and inclusion (Knobloch et al., 2011).
The identification of studies
At the outset, sources on AI in the public sector were retrieved from the Web of Science (WoS) database, the
most prominent scholarly literature pool worldwide. The search strings were restricted to the key concepts, such
as “artificial intelligence”, “AI”, “challenges”, “obstacles”, “public services”, “service delivery”, “public
sector”, “public administration”, and “South Africa”. Considering that cutting-edge technologies (e.g., AI) are
volatile and rapidly evolving due to the ever-changing business landscape (political, economic, social,
technological, and ecological – PESTE) in the digital era, this study only included current sources from the 2023
to 2025 academic years. This enabled the authors to access current debates on AI.
The screening process
Following a comprehensive search of the literature in WoS, the screening process has been initiated. The sources
were meticulously assessed by scrutinising the abstracts to ascertain their suitability for this study. Aside from
that, the sources were critically and systematically synthesised to discover pertinent themes and patterns. While
themes and patterns related to the subject were identified during the screening process of the sources, the authors
observed alignment with the study’s purpose. Three autonomous evaluators were delegated to assess the sources
to circumvent prejudice during the selection procedure and improve data reliability when scrutinising the
sources. Slight differences of views were reflected and settled, finally reaching a consensus. The EndNote
Reference Manager (Version 21.5) was utilised to efficiently manage and handle the replicated references and
create a complete data file.
Considering the eligibility and inclusion criteria
After thoroughly screening the academic sources, the authors considered the eligibility criteria (inclusion and
exclusion). Only studies published in the public sector context were included to enhance reliability and validity.
To this end, however, apart from the irrelevant sources, non-English studies were excluded due to the complexity
of interpretation. Table 1 below underscores the inclusion and exclusion criteria of this study.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 1372
www.rsisinternational.org
Table 1 Eligibility requirements
Item Description
Domain Only sources published within the public sector domain were included in the study.
Literature This study included peer-reviewed journal articles and conference papers.
Key concepts The search for key concepts was limited to the following terms: “artificial intelligence”,
“AI”, “challenges”, “obstacles”, “public services”, “service delivery”, “public sector”,
“public administration”, and “South Africa”.
Period The study included sources published between 2023 and 2025 to ensure the prevalence of
the academic discourses.
Research methods Studies that included both qualitative and quantitative research methods were considered.
Language Only sources written in English were included. Sources in other languages were excluded
based on translators' difficulty accessing them.
Data quality evaluation
Following the screening process, the data were systematically and thoroughly assessed to determine their value
and enhance quality assurance, while also ensuring the reliability and validity of the sources. Furthermore, the
rigorous quality appraisal procedure helped identify sources that could significantly contribute to the study in
response to the research question, thereby achieving the study's purpose. Most importantly, to improve the
quality of sources consulted, only research published in the most renowned journals was deemed suitable to meet
the quality appraisal requirements. The data quality evaluation (DQE) was based on the factors underscored in
Table 2.
Table 2 Data quality evaluation requirements of eligibility
DQE Code Specification Response
DQE 1 Is the primary purpose/objective explicitly stated? Absolutely = 1,
Moderately = 0,5
and
Not applicable = 0
DQE 2 Is the theoretical framework (if applicable) clearly explained and aligned
to the study?
DQE 3 Is the underpinning methodological choice thoroughly elucidated and
rationalised?
DQE 4 Do the findings offer explicit implications and suggestions?
DQE 5 Does the research attain secondary objectives, if any?
RESULTS
The comprehensive search of the literature using the WoS search engine yielded 1003 sources (as illustrated in
Figure 2), in addition to 7 sources suggested by the experts. While the authors considered the eligibility
requirements, the sources without abstracts, non-English, identical and irrelevant, amounting to 689, were
eliminated, leaving 314. Of the 314 sources, 243 full-texts were removed based on the study's eligibility
requirements, leaving 71. Furthermore, the quality evaluation was conducted on 71 sources, resulting in 20
sources being included in the systematic analysis. Most notably, only sources that focused on AI in the public
sector were suitable for inclusion in the broad review, while also considering the South African studies that
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 1373
www.rsisinternational.org
informed the contextual factors influencing the adoption of AI. Ultimately, 20 sources, comprising 16 peer-
reviewed journal articles and four conference papers, were included in the systematic analysis. The statistical
contribution of sources is as follows: 16 (80%) of peer-reviewed journal articles and four (20%) of conference
papers.
Figure 2 PRISMA flow diagram (Page et al. 2021)
DISCUSSION OF FINDINGS
The comprehensive and rigorous systematic literature analysis was undertaken in response to the research
question: What are the key challenges faced by the South African public sector in adopting AI? This section
elucidates the implications of the themes derived from the PRISMA technique, such as digital infrastructural
deficit, digital illiteracy, policy and regulatory gaps, and ethical dilemmas.
Digital infrastructural deficit
Generally, although digital infrastructure plays an integral role in driving digital innovation (including AI),
accelerating public services and boosting economic expansion of the country, most African countries are faced
with limited digital infrastructure (e.g., broadband internet connectivity), encumbering the accomplishment of
digital solutions in the digital arena (Osei, 2024). To this end, Osei’s study confirms a positive linkage between
digital infrastructure and innovation. Like other emerging countries, Naidoo (2024) corroborates that since South
Africa faces funding challenges concerning digital infrastructure, particularly in remote areas affecting
designated (marginalised) groups, this further exacerbates the persistent socio-economic discrepancies (digital
divide in those communities and the application of AI in those communities. This is consistent with the ideas of
Zindi and Majam (2024), who opine that while digital infrastructure (including electricity distribution) is one of
the fundamental impediments to downplaying the development of AI and the provision of digitalised service
delivery by municipalities in South Africa, the effective and efficient public services have been drastically
disrupted. That said, it is worth remarking that digital infrastructure is a crucial aspect not only in enhancing AI
adoption but also in the uptake of other emerging technologies such as cloud computing, big data, virtual
platforms, and blockchain in the context of the public sector (Henk & Henk, 2025; Modiba, 2025).
Digital illiteracy
The public sector, specifically in developing countries, is characterised by digital illiteracy or the absence of
skilled professionals compared to the private sector (Mahusin et al., 2024). The lack of investment in
comprehensive training and development of public servants aggravates this issue. Furthermore, the scarcity of
digital competencies in South Africa has been identified as a significant barrier to the adoption of AI tools and
techniques (e.g., machine learning) in the public sector (Baloyi, 2025; Chilunjika, 2024). More to this,
notwithstanding AI being regarded as a vehicle for fast-tracking seamless public services and ensuring improved
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 1374
www.rsisinternational.org
productivity, it is contended that most public servants across the South African public sector (including local
government) lack appropriate competencies to operate AI-related technologies in the digital world (Motadi,
2024; Zindi and Majam, 2024). It is argued that apart from the critical drawbacks brought about by AI, such as
job displacement and skepticism, employees in the public sector are negatively impacted by the absence of
expertise necessary to drive AI-enabled technologies (Henk & Henk, 2025). Most remarkably, older public
servants seem reluctant to accept modernised public services (i.e., the integration of AI into processes) imposed
by technological changes, primarily due to a lack of interest in acclimating and adjusting to the new processes
(Barodi & Lalaoui, 2025). That being the case, it drags the progress of fulfilling modernised public services as
the traditional paradigm supersedes AI applications.
Policy and regulatory gaps
Policy and regulatory frameworks are the focal point of AI acceptance in the public sector. To this effect, many
developing countries are plagued by the scarcity of policy and legislative models guiding the development of AI
technologies (Motadi, 2024), as well as AI-related legal uncertainties that often hinder the smooth deployment
of ethical AI – specifically, concerns over data security, privacy, and transparency (Androniceanu, 2024). Even
though the South African government has taken radical steps to assimilate AI tools into systems in diverse areas
in the public sector, include but not limited to, law and judiciary, agriculture, social services, home affairs,
education, and health, inter alia, the regulatory framework governing AI is still blurred (Janneker, 2025;
Chilunjika, 2024; Motadi, 2024). For instance, regardless of the attempts made to establish AI policy (e.g.,
“South Africa National Artificial Intelligence Policy Framework”), the framework remains a draft and is
nevertheless in the process of endorsement by the Minister of Communication and Digital Technologies after a
thorough consultation with the public and different stakeholders (Baloyi, 2025). In a similar vein, it is vital to
develop robust AI legal mechanisms and enforcement to circumvent potential ethical concerns (such as public
trust, fairness, and bias), thereby enhancing transparency in AI algorithms (Tomazevic et al., 2024). This may
help minimise bias in AI usage. Furthermore, as AI rapidly penetrates public administration due to volatile
business conditions, developing binding laws and regulations has become mandatory to accelerate the delivery
of cost-effective services (Henk & Henk, 2025; Madan & Ashok, 2023).
Ethical dilemmas
Ethical concerns are a crucial consideration when adopting AI technologies in the public sector. For instance,
moral concerns, including but not limited to data security, privacy, transparency, accountability, and the bias of
algorithms, play a prominent role in warranting the ethical use of AI, not only in providing efficient public
services but also in enhancing fairness and non-discrimination for its users (Shekgola & Modiba, 2025).
Additionally, while AI technologies present potential gains (e.g., economic growth, cost savings, and enhanced
decision-making) through automated decision-making, especially in advanced economies globally, developing
countries like South Africa and other African countries are overwhelmed with ethical conundrums, hindering
the utilisation of AI by citizens (Baloyi, 2025). It is substantiated that the persistent ethical dilemmas emphasised
earlier lead to AI initiatives’ failures and sluggish service delivery, thus disrupting the achievement of the public
sector’s strategic objectives and the NDP’s SDGs goals (Naidoo, 2024), at the same time, eroding public trust
(Maleka & Maidi, 2024). It is important to integrate ethical principles (data security and privacy) and governance
standards (accountability and transparency) when designing AI legal frameworks and models, and in developing
and implementing AI systems to prevent unnecessary AI resistance by citizens and other stakeholders.
Despite the challenges delineated, the South African government endeavours to embrace AI-driven technologies
in the public sector. For instance, Stellenbosch University in South Africa has compiled policy data related to
the global application of AI. The University found that AI is prevalent in providing public services across various
sectors by the South African government (Stellenbosch University, 2025). Most profoundly, the policy identified
AI tools that governments are currently employing or have yet to utilise in rendering services. The following
serves as an example.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 1375
www.rsisinternational.org
Table 3 AI tools in the South African public sector (The Stellenbosch University, 2025)
No. AI Type of service Sector Status
1. GRIT-GBV Zuzi
chatbot.
Public order and safety. Law and Judiciary. Implemented.
2. Project Vikela. Discover illicit rhino poaching. Agriculture. In development.
3. Automated Biometric
Identification System.
AI-powered matching of fingerprints,
faces & palm prints.
Home Affairs. Operational.
4. SARS “AI Assistant”
chatbot.
Natural-language bot answers tax,
customs & traveller queries.
South African
Revenue Service
(SARS).
Operational.
5. AI-assisted TB
screening vans
(CAD4TB/Qure.ai).
Portable X-ray units score images
with AI.
Health. Pilot.
6. GovSearch. An intelligent search engine
centralises reports, policies, and
sector plans.
Cooperative
Governance &
Traditional Affairs.
Planned (not yet
implemented).
7. CSIR Body and number
plate recognition
system.
An integrated face, body and license
plate recognition system for secure
access control and surveillance,
combining multiple biometric
technologies for enhanced security.
Council for
Scientific and
Industrial Research
(CSIR).
Pilot/Tested.
Given the disparate implementation status highlighted above, it is evident that South Africa is in a nascent phase
of AI adoption. Although the South African government’s AI initiatives in service delivery efforts are necessary,
as indicated in Table 3, further efforts are needed to enhance AI adoption in the public sector. Yet again, as the
South African public sector continues to adopt various AI-enabled technologies to streamline internal processes
and augment efficiency, its services to citizens can be significantly improved.
CONCLUSION
This study aimed to investigate the key challenges encountered by the South African public sector in adopting
AI. Even though AI has been praised by multiple governments around the world for its efficiency in public
services, the South African public sector has yet to reach that level. The identified themes (i.e., digital
infrastructural deficit, digital illiteracy, policy and regulatory gaps, and ethical dilemmas) play an adverse role
in realising AI initiatives, although they are imperative in remedying the status quo and assisting in expediting
the South African public sector's adoption of AI-driven technologies. It is suggested that long-term investment
in infrastructure development, digital skills, and funding for digital technologies can be considered key to
harnessing and nurturing AI technologies in the South African public sector. furthermore, whereas the global
business environment is ever-evolving and AI is continuously becoming part of our daily lives, the rate at which
various sectors (public and private) embrace AI is significant. As such, the South African public sector cannot
disregard the adoption of AI technologies if modernised and better ways are to be found to improve the efficiency
and streamlining of public services.
This study contributes to the body of knowledge by exploring the key challenges faced by the South African
public sector in adopting AI. Addressing these key challenges is an essential aspect that can enable the South
African public sector to thrive in the digital era. Furthermore, policymakers, practitioners, and decision-makers
in the public sector can consider the findings of this study to be imperative for policy development and in
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 1376
www.rsisinternational.org
remodelling internal processes through selected digital strategies, specifically for those in the early stages of
adopting AI-driven technologies.
Conflict Of Interest
The author declares that he has no financial or personal relationships that may have inappropriately influenced
him in writing this article.
Statement Of Ethical Approval
The present research work does not contain any studies performed on animal/human subjects.
Data Availability
The authors confirm that the data supporting the findings of this study are available within the article and in the
bibliography.
REFERENCES
1. Androniceanu, A. (2024). Artificial intelligence in administration and public management. Administratie
si Management Public, 42, 99-114.
2. Awa, H. O., Ojiabo, O. U., & Orokor, L. E. (2017). Integrated technology-organisation-environment
(TOE) taxonomies for technology adoption. Journal of Enterprise Information Management, 30(6), 893-
921.
3. Baker, J. 2012. The technology–organisation–environment framework. Information Systems Theory:
Explaining and Predicting Our Digital Society, 1: 231-245.
4. Baloyi, W.M. (2025). A systematic analysis of ethical and governance concerns relating to artificial
intelligence adoption in the South African public sector. International Journal of Research and Innovation
in Social Science, IX(IX), 3177-3190.
5. Baloyi, W.M., & Beyers, L.J.E. (2019). The impact of financial malpractice on service delivery: A
conceptual examination of literature, in 4th Annual International Conference on Public Administration
and Development Alternatives, 03–05 July, Johannesburg.
6. Barodi, M., & Lalaoui, S. (2025). Civil servants’ readiness for AI adoption: The role of change
management in Morocco’s public sector. Problems and Perspectives in Management, 23(1), 63-75.
7. Bryan, J. D., & Zuva, T. (2021). A review on TAM and TOE framework progression and how these
models integrate. Advances in Science, Technology and Engineering Systems Journal, 6 (3), 137-145.
8. Bester, J. (2024). Exploring the preparedness of South African rural municipalities in the adoption and
use of artificial intelligence to improve service delivery. Journal of Public Administration and
Development Alternatives, 9(1), 27-41.
9. Chilunjika, A. (2024). A review of the risks, challenges and benefits of using artificial intelligence (AI)
technologies in public policy-making in South Africa. Social Sciences, Humanities and Education
Journal, 5(3), 393-411.
10. Henk, A. & Henk, O. (2025). From antecedents to outcomes: A structured literature review on AI
implementation in public sector organisations. Proceedings of the 58th Hawaii International Conference
on System Sciences. https://doi.org/10.24251/hicss.2025.228.
11. Janneker, W. (2025). The role of AI in transforming healthcare in South Africa. SAMJ: South African
Medical Journal, 115(5B), 66-69.
12. Knobloch, K., Yoon, U., & Vogt, P. M. (2011). Preferred reporting items for systematic reviews and
meta-analyses (PRISMA) statement and publication bias. Journal of Cranio-Maxillofacial Surgery,
39(2), 91-92.
13. Madan, R., & Ashok, M. (2023). AI adoption and diffusion in public administration: A systematic
literature review and future research agenda. Government Information Quarterly, 40(1), 1-18.
14. Mahusin, N., Sallehudin, H., & Satar, N. S. M. (2024). Malaysia’s public sector challenges of
implementation of artificial intelligence (AI), in IEEE Access, vol. 12, pp. 121035-121051,
https://doi.org/10.1109/access.2024.3448311.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue X October 2025
Page 1377
www.rsisinternational.org
15. Maleka, S., & Maidi, C. (2024). The societal implications of technological innovations and AI in South
Africa. Public Sector Conference, held at Cape Town, 15 April 2024, pages 1-22.
16. Malope, L. J. (2025). Unlocking the potential: Exploring the influence of artificial intelligence on e-
health enhancement in South Africa. Journal of Economic and Social Development (JESD)–Resilient
Society, 12(1), 146-156.
17. Modiba, M. (2025). Adoption of artificial intelligence to enhance records management practices at the
Gauteng Department of Education in South Africa. Collection and Curation, 44(1), 9-17.
18. Motadi, M. S. (2024). Challenges and Opportunities: The Role of Artificial Intelligence in Reinventing
Public Administration in South Africa. International Journal of Public Administration in the Digital Age
(IJPADA), 11(1), 1-20.
19. Naidoo, G. M. (2024). The potential of artificial intelligence in South African rural development. OIDA
International Journal of Sustainable Development, 17(11), 207-218.
20. Osei, D. B. (2024). Digital infrastructure and innovation in Africa: Does human capital mediates the
effect?. Telematics and Informatics, 89, 1-13.
21. Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., ... & Moher,
D. (2021). The PRISMA 2020 statement: An updated guideline for reporting systematic reviews.
Research Methods and Reporting, 10(89), 1-11.
22. Saal, P. E., Chetty, K., Ntshayintshayi, N., Moosa, T., & Masuku, N. (2025). A scoping review of the
integration of artificial intelligence in primary and secondary schools from 2020 to 2024: Policy
implications for South Africa. Journal of Education (University of KwaZulu-Natal), 98, 62-85.
23. Shekgola, M., & Modiba, M. (2025). Utilising an AI chatbot to support smart digital government for
Society 5.0 in South Africa. South African Journal of Information Management, 27(1), 1-10.
24. Sohrabi, C., Franchi, T., Mathew, G., Kerwan, A., Nicola, M., Griffin, M., ... & Agha, R. (2021).
PRISMA 2020 statement: What's new and the importance of reporting guidelines. International Journal
of Surgery, 88, 1-4.
25. The Stellenbosch University. (2025). Catalogue of AI tools in government. Policy Innovation Lab.
Available at: https://policyinnovationlab.sun.ac.za/tool/catalogue-of-ai-tools-in-government/ (Accessed
20 September 2025).
26. Tomazevic, N., Murko, E., & Aristovnik, A. 2024. Organisational enablers of artificial intelligence
adoption in public institutions: A systematic literature review. Central European Public Administration
Review, 22(1), 109-138.
27. Tornatzky, L. G. & Fleischer, M. (1990). Processes of technological innovation. Lexington, MA:
Lexington Books.
28. Tveita, L. J., & Hustad, E. (2025). Benefits and challenges of artificial intelligence in the public sector:
A literature review. Procedia Computer Science, 256, 222-229.
29. Zindi, B., & Majam, T. (2024). Leadership in the Age of Digital Transformation: Building Capacities for
Public Servants in South African Municipalities. Journal of Public Administration, 59(3), 379-397.