1 Introduction

To change classroom dynamics and encourage student-centered learning, several educators have embraced technology integration for educational purposes [39]. One of the trending technologies is artificial intelligence. Artificial intelligence (AI) has evolved unfathomably in the twenty-first century, as it is widely used across many human endeavours for various purposes. It is a broad concept that includes various computational tools and methodologies that seek to replicate human cognitive functions and enable machines to perform tasks that usually require human intelligence [47]. Artificial intelligence can access and explore every facet of the digital world that people operate within [61]. This approach is not limited to customized healthcare service delivery, safety measures, shopping, smart homes, or many other services, as the education sector must be included [24, 57]. A rising variety of devices, using methods such as modern deep learning, could deliver intelligent services by inferring or behaving like people, as AI could replace educators and support staff in schools (Al Ka’bi, [6]). The deployment of AI in numerous sectors is projected to increase operational efficiency, lower costs, and improve users’ experiences [91]. Artificial intelligence is revolutionizing all facets of human existence, work, and education. AI is used in the educational sector for several purposes and by different individuals. Academics responsible for teaching undergraduates and conducting landmark research may not be left behind in using AI. Their goals for using AI may not be limited to their core responsibilities,they could also diverge from the goals for which undergraduates use AI.

Science, Technology, Engineering, and Mathematics (STEM) contribute massively to the development of any nation. Its education has been prioritized by both developed and developing nations [60]. STEM has aided the development of AI tools and the usage of the AI tools has been beneficial to the advancement STEM. The usage of AI tools in higher education, especially in STEM fields, has gained the attention of researchers in recent years. Despite the capabilities of AI-powered tools and applications to improve administrative procedures, enhance teaching and learning environments, and encourage academic and student creativity and teamwork, its acceptance and usage in higher education are not even [72]. Some nations or institutions fall behind in their attempts at AI strategic adoption [25, 28]. The African Higher Education Institution (AHEI) appears to be one of the understudied in the usage of AI tools. Though STEM education is becoming more important day after day, and AI transforming power is popularly acknowledged, the current state, challenges, and strategic goals surrounding its integration among STEM academics and undergraduates in African higher education institutions are not fully explored. Through the review of the literature, this study aims to investigate the goals of academics and undergraduates in African higher education for integrating AI. This systematic review examines the strategic goals of AI integration in STEM education across African higher institutions. By uncovering the state of AI adoption, peculiar challenges, and institutional goals, the review offers invaluable insights to facilitate the effective implementation of transformative AI technologies. Furthermore, the review's focus on the underrepresented African context contributes to a more diverse and inclusive understanding of the challenges and opportunities surrounding AI integration in STEM higher education. This will foster the advancement of educational practices in African higher institutions. When the goals that motivate academics and undergraduates to integrate AI into African higher education are identified, this will foster innovative policy decisions, resource allocation, and curriculum development. Additionally, this study provides insight into fostering innovative AI development, enhancing research capabilities, and promoting inclusiveness in STEM education, contributing to the primary goal of socioeconomic growth and technological advancement in Africa.

2 Literature review

The utilization goals of AI at African higher educational levels are the focus of this literature review. This review investigated AI integration’s importance and potential benefits in African higher education. The potential benefits of adopting AI in African higher education involve the core activities of teaching, learning, and research carried out in African higher education institutions. The benefits encompass personalized learning, improved learning outcomes, enhanced efficiency and productivity, adaptive assessment and evaluation, intelligent student support, predictive analytics, decision-making, scalability, and cost-effectiveness. The goals for AI utilization vary between academics and undergraduates. The review emphasizes that academics in African higher education are increasingly utilizing AI to enhance research, generate academic content, substitute traditional tools, assess and grade, enhance teaching, and complete administrative tasks. However, there is a need for additional research on the scope of and variations in responsible and effective AI integration across disciplines and challenges and best practices to increase academic productivity, teaching quality, and student learning outcomes in African higher education.

Conversely, the review showed that undergraduates predominantly use AI to facilitate note-making, efficiently complete assignments, generate academic content, substitute traditional tools, and enhance learning. They leverage AI to solve academic problems more accurately and with greater seamlessness. More research needs to be done on the potential benefits, challenges, and best practices of responsible AI integration in African higher education to help students learn and succeed.

3 AI integration in African higher education

In Africa, AI is permeating a growing number of processes and solutions. Many locally created AI solutions support people's decision-making in several vital areas, fostering the development of young data scientists in these areas [3, 81]. Recently, Africa has emerged as one of the global leaders in embracing technology [13, 62], driven in part by the necessity of adapting and overcoming the unprecedented challenges imposed by the COVID-19 pandemic [70]. Despite concerns about their use, educational institutions seem to be increasingly adopting AI-driven technologies. How academic programs are conducted in higher institutions has evolved significantly since the emergence of AI, especially in the post-COVID era [10]. Studies highlight the potential of AI in transforming higher education in Africa. AI can enhance student learning outcomes, promote collaborative and personalized learning experiences, and contribute to achieving Sustainable Development Goals [65]. The integration of AI technologies in African universities can foster innovation, industrialization, and economic development [92]. Studies have investigated the integration of AI in AHEIs and have shown that AI technologies have been adopted for various roles but has not been fully integrated into the higher education system. AI has been integrated in AHEIs for teaching and learning purpose [4], enhancement of academic achievement [29], personalized learning [12], academic engagement [15] among others. However, various challenges hinder the full adoption of AI in African higher education [41].

One of the most promising areas of AI integration in African higher education is the development of intelligent tutoring systems (ITS). These adaptive learning platforms leverage AI algorithms, natural language processing and machine learning to create to provide personalized instruction and feedback to students [23]. For example, the website “Knewton” (www.knewton.com) and “Century Tech” (www.century.tech) platform offers an AI-powered learning platform that analyzes student performance data to deliver customized lessons and assessments. Another substantial application of AI in African higher education is the use of automated feedback and grading systems. The website "Gradescope" (www.gradescope.com) utilizes computer vision and machine learning to streamline the grading process, allowing instructors to provide detailed feedback and assess student work efficiently [53]. Additionally, the “Quillionz” (www.quillionz.com) platform employs natural language processing to generate personalized feedback on student writing, helping them improve their skills [64]. AI-powered tools are also being integrated into the administrative aspects of higher education, such as scheduling and classroom management. The website “Edwiser” (www.edwiser.org) offers an AI-driven scheduling system that optimizes class timetables and resource allocation, reducing conflicts and improving utilization [87]. Similarly, the “ClassDojo” (www.classdojo.com) platform uses AI to facilitate classroom management, enabling instructors to track student engagement, communicate with parents, and provide real-time feedback.

4 Challenges of AI integration in African higher education

Integrating Artificial Intelligence (AI) in African higher education has different opportunities and challenges. AI has the potential to transform and resolve issues associated with various activities being carried out in African Higher Education through personalize learning experiences, enhance student engagement, and improve administrative efficiency [19, 27]. However, significant challenges exist, including underdevelopment, competency gap, lack of quality education, and inadequate 21st century skills among graduates [49, 52]. Integrating AI in education could foster innovation and industrialization in Africa, but it requires addressing issues such as the digital divide, data privacy, and potential biases in AI algorithms [27, 36]. Several AHEIs are not prepared to integrate AI, as this is evident in their lack of policy presentation that could guide its usage among the members of the institutions [2]. Many of these institutions are still unsure of how ethical or professionally acceptable AI tools are. This has left many AI tool users especially undergraduates in AHEIs apprehensive about the acceptability of its usage [30]. Ethical considerations are crucial, particularly regarding data collection and analysis of student information [27]. Issues surrounding ethical usage have been on the frontline among the challenges confronting the integration of AI in AHEIs [20]. Another challenge is the limited infrastructure and technological resources available in many parts of the continent [71]. Many universities lack access to reliable high-speed internet, adequate computing power, and specialized software or hardware required for advanced AI applications. This infrastructure gap hinders the ability of these institutions to fully leverage AI tools and limits the potential for widespread adoption [74]. Another key challenge is the limited technological literacy and digital skills among academics and African higher education students [7]. Extensive training and capacity-building programs are needed to equip academics and students with the necessary knowledge and skills to effectively utilize AI-powered tools and applications. A lack of familiarity with AI and its potential benefits can create resistance to adoption and integration within the academic community [90]. Addressing these technological literacy gaps is crucial for fostering a supportive environment for AI integration in STEM education. To integrate AI in African higher education successfully, collaborative efforts involving governments, educational institutions, and technology developers are necessary, along with policies that promote gender equity, cultural diversity, and labour market adaptability [19, 36].

5 Importance and benefits of AI integration in higher education

Artificial intelligence (AI) is becoming a more popular tool used for teaching and learning [33]. Since the closure of schools due to the 2019 coronavirus pandemic, artificial intelligence has gained increased importance in overcoming distance barriers in teaching and learning (Darayseh, [5]). In higher education, the benefits of AI are most pronounced in teaching, learning, and research, as these represent the core activities undertaken in higher education institutions. The adoption of AI in higher education has several benefits, including personalized learning [22], improved learning outcomes [68], enhanced efficiency and productivity [73], adaptive assessment and evaluation [42], intelligent student support [22, 57], predictive analytics and decision-making [85], scalability and cost-effectiveness [51], continuous improvement and innovation [54], and research enhancement [40].

According to Surugiu et al. [79], AI is beneficial for carrying out administrative tasks by saving time when completing much work within the shortest possible time. Personalized learning is one of the key composite benefits of AI for individual learners [22, 46, 88]. Students are more likely to be motivated, engaged, and independent when they use technologies that are based on AI, which provides immersive and personalized learning experiences [14, 63]. This indicates that, with AI-driven instruction, undergraduates can learn at their own pace, potentially addressing challenges associated with individual learning patterns. AI has opportunities for improving both the quality of instruction provided by educators and students' learning results when integrated into the academic system [33, 66]. According to Owen et al. (2023), AI has proven to be efficient when assessing and providing feedback. AI provides an intelligent tutoring system (ITS), an automated grading system (AGS), and predictive analytics. These tools provide students with personalized help and feedback, catering to the unique learning styles of each student and thus improving learning outcomes [44, 57, 68]. AI tutoring systems provide students with individualized support, guidance, and assessment by customizing learning materials according to their unique learning styles or levels of expertise [44]. Using AGS, educators can grade students efficiently while simultaneously streamlining the process,students receive feedback, and their writing abilities improve [42, 68]. This efficiency allows educators to focus on higher-value tasks, making them feel more productive. Students who are at risk may be identified via the use of predictive analytics, which then enables targeted interventions and assistance to be provided to enhance their academic performance [68, 85]. According to Goel and Polepeddi [35], educators may repurpose their time to perform higher-value tasks when using artificial intelligence (AI) to answer students' repetitive, simple queries in online discussion forums. Using AI in teaching and learning allows for decoding learners' performance, growth, and potential via their clickstream data [76]. The review showed that integrating AI in higher education is becoming increasingly important and beneficial, with key advantages in personalizing learning, enhancing teaching and learning outcomes, improving administrative and assessment processes, providing intelligent student support, and enabling data-driven decision-making. Thus, there is a growing need to examine further the opportunities for the responsible and effective adoption of AI technologies in the academic ecosystem.

6 Goals for AI Integration in Higher Education

AI can transform educational institutions by personalizing teaching methods to meet individual student needs, providing timely feedback, executing administrative tasks, assisting with grading and assessment, and allowing educators to focus on curriculum development and quality instruction [77]. According to Surugiu et al. [79], AI can support the core tasks of teaching, learning, and research in various ways. The primary stakeholders in higher education are academics and students, typically referred to as undergraduates. Academics and undergraduates use AI for different purposes [10, 33, 40, 67]. The goals for AI used by these key players may differ, underscoring the importance of examining these goals from the perspectives of academics and undergraduates.

6.1 Goals of using AI by academics

Academics have numerous tasks to accomplish before, during, and after teaching sessions and research activities. AI can significantly aid academics in these endeavours. Several studies have indicated the various purposes of academics adopting AI, some of which include enhanced research [56], generating academic content [15, 75], substituting traditional teaching tools [32, 86], assessing and grading [31, 43], enhancing teaching [82], and completing administrative tasks [34]. Academics are integrating AI to improve the accuracy and speed of data analysis, information retrieval, and informed decision-making [17, 26]. In higher education, academics are tasked with preparing academic content, such as lecture notes, to facilitate teaching and provide students with materials for assessment preparation, AI adoption enhances these processes by enabling efficient and accurate content creation, thereby improving the quality and speed of content generation to support teaching and learning activities [57]. Academics utilize AI as a beneficial alternative to traditional research tools, such as Google, in information gathering, enabling greater efficiency and accuracy in research procedures through sophisticated automation and data processing capabilities [32, 86]. Academics utilize AI technologies that offer automated assessment and grading, providing efficient and consistent evaluation techniques that save time and ensure fair and objective grading standards [43, 68, 85]. In addition, AI integration in higher education increases teaching efficiency by customizing learning experiences, delivering real-time feedback to students, and providing creative teaching tools that adapt to varied learning styles and demands [18, 35, 38]. Several academics combine administrative work with their core responsibilities,AI is used to improve tasks, such as scheduling, grading, and reporting, enhancing inclusive efficiency and enabling them to concentrate on more meaningful interactions with students [79]. The review revealed that academics aim to use AI to ease the rigour of teaching, notetaking, assessment, and the research process.

6.2 Goals of using AI by undergraduates

The use of AI by undergraduates appears to be intended to help them solve academic problems seamlessly and more accurately. These academic problems vary from one undergraduate to another. Studies have revealed that undergraduates use AI-driven tools for facilitating notetaking [59], efficient assignment completion [21], generating academic content [58], substitute for traditional tools [48], and enhanced learning [15]. During lectures, undergraduates use various AI tools to take notes more quickly and precisely [39, 45]. For example, undergraduates make voice notes with a technological device and later use AI tools to transcribe or convert them to text.

Moreover, some use AI tools to convert text to audio files that people listen to in their leisure time [37]. According to Joshi et al. [50], undergraduates increasingly depend on AI to complete assignments effectively and precisely. This could be a way of saving time and reducing stress by generating solutions to assignments. As evident in Selim’s study (2024), undergraduates use AI to develop academic content more precisely. Moreover, undergraduates who carry out research tasks such as final year projects or seminars depend on AI to enhance their research processes [80]. This approach might enable them to access various resources and information more efficiently and accurately. Thus, AI may replace conventional search engines such as Google, enabling undergraduates to obtain desired information quickly and more accurately [11, 55]. Personalized learning is one of the core benefits of AI-driven educational tools, and undergraduates can improve their learning experiences through access to personalized materials and feedback on the academic tasks completed [83]. The review revealed that undergraduates employ AI for various purposes, primarily directed toward academic problem solving, including assignments, notetaking, information retrieval, and completing their final year projects.

7 Objective of the study

The study aims to comprehensively evaluate the literature published over the last ten years on the goals of adopting artificial intelligence (AI) in African higher education, focusing on academics and undergraduates in STEM education. This study specifically examines the strategic goals of utilizing AI tools by academics in STEM education and undergraduates' usage goals for adopting AI tools in STEM education.

8 Research questions

  1. 1.

    What are the strategic goals of STEM higher education academics in Africa for utilizing AI tools?

  2. 2.

    What are the goals for African higher education undergraduates' adoption of AI tools in STEM education?

  3. 3.

    What AI tools are commonly used by academics and undergraduates in African higher education institutions?

  4. 4.

    How are academics and undergraduates benefiting from adopting AI, and what are the possible challenges faced in STEM education in African higher education institutions?

9 Methodology

This was a systematic review in which recent studies from 2015 to 2024 (10 years) that focused on the specific goals of adopting AI tools by academics and undergraduates in African higher education were examined. The authors chose 2015 as the starting point for their systematic review to capture the emerging trends and developments in AI integration within African higher education before the COVID-19 pandemic, which accelerated the widespread adoption and accessibility of various AI tools and technologies across educational institutions. It is also important to note that the 2024 articles included were published in the year's first half. Three research questions were raised to guide the study. This systematic review was carried out according to the principles of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement [69]. The PRISMA 2020 guidelines were used to identify and incorporate prior research findings and gather data to address the research questions. The study was carried out in five phases, as represented and summarized in Fig. 1.

Fig. 1
figure 1

Phases of the systematic review

Full size image

9.1 Search strategy

After defining the scope of the study, focusing on AI usage/adoption in African higher education by academics and undergraduates, the following research questions were raised in accordance with the scope of the study. The keyword search was carried out with standard academic databases, with a primary focus on the Scopus (https://www.scopus.com/search/form.uri?display=basic#basic) and Semantic (https://www.semanticscholar.org/) databases, which are renowned for their comprehensive coverage of academic literature. The results were subsequently analyzed with Google Scholar and Crossref registries.

In an attempt to identify a wide range of articles, some keywords were combined. This strategy helped verify the literature for synonyms. To perform the search, the Boolean operators ‘AND’ and ‘OR’ were used to combine the items in the search. The search was structured as follows:

9.1.1 (“Artificial intelligence”) AND (“undergraduate” OR “university students” “Academics” OR “educator” OR “lecturer”).

The keyword artificial intelligence was combined with undergraduate or university students or ‘academics or educator’. This approach enables the visibility of articles focusing on artificial intelligence usage or adoption by undergraduates and academics. Using keywords such as university students and educators provides insight into studies where academics are referred to as lecturers or educators and undergraduates are referred to as university students or students.

9.2 Inclusion and exclusion criteria

The search of the Scopus database was limited to studies published between 2015 and 2024 (10 years), journal articles, and conference proceedings. The decision to exclude non-peer-reviewed publication types was based on the higher quality standards and rigor of the peer review process for journal articles, which is essential for providing reliable empirical evidence to inform AI integration policies and practices in African higher education. It was observed that some of the articles retrieved from the Scopus database were also available in the Semantic database, and such articles were not added. Table 1 summarizes the criteria for including and excluding articles considered in this review.

Table 1 Summary of the exclusion and inclusion criteria
Full size table

A total of 1471 journal articles and conference proceedings were identified from the databases; 1226 articles were excluded, 56 duplicates were excluded, 1106 were ineligible and removed by automation, and 64 were identified manually to be ineligible from the topics that were out of context. Next, 245 articles were further screened (57 from Scopus, 116 from Semantic, and 72 from other databases); 178 were found to be ineligible and subsequently excluded. A total of 67 articles were sought for retrieval. After examining the abstracts of these articles, 39 articles were not retrieved because they could not meet the criteria for consideration. Overall, 28 articles were fully accessed and retrieved, 16 of which were further excluded; six of these articles neither focused on the usage of AI nor had samples from African higher education institutions; two reports were nonempirical; and eight did not focus on any of the STEM disciplines. Overall, 12 articles were considered for the review. Figure 2 shows the PRISMA flowchart for the inclusion and exclusion of reports.

Fig. 2
figure 2

PRISMA flowchart for inclusion and exclusion of reports

Full size image

10 Coding

The 12 articles considered in the study were coded using both deductive and inductive coding techniques to address research questions. Deductive coding involves using pre-determined, a priori codes to classify the data, while inductive coding allows the codes and themes to emerge directly from the data through a grounded coding approach [22, 89]. The pre-existing codes used in the deductive coding includes years, countries, Africa higher education, author affiliations, academics, undergraduates and fields. The main goal of this study is to determine the strategic goals of using AI tools among academics and undergraduates in African higher education, the grounded coding technique was used to enable the identification of patterns in AI tools usage among academic and undergraduates in Africa higher education be identified via the data [22, 78]. To understand how AI tool was being used in higher education, an inductive, grounded theory approach was employed. Researchers extracted information from the articles on the purpose of using AI tools by academics and undergraduates, the forms of AI tools utilized, as well as the benefits and challenges of adopting these tools. To accurately capture the terminology used by the authors and ensure consistency with their findings, the researchers employed in vivo coding, which uses the language directly from the research articles. The constant comparative method was used in the grounded coding process, allowing important themes and categories to emerge iteratively from the data, rather than fitting the data to pre-existing frameworks.

11 Results

Table 2 and Fig. 3 indicate that the highest report 33.3% came from South Africa [9, 15, 65, 84], 16.7% of the reports came from Morocco [12, 16], while Lesotho [8], Egypt [1], Zambia [53], Nigeria [64] and Ghana [29] had a lower turnout of reports representing 8.3%. Additionally, a multinational [87] with 22 African countries involved in the report was included. The turnout rate could symbolise how popular or acceptable AI technologies are in the countries considered. This could also indicate that there are few studies on the use of AI by STEM undergraduates and academics in African higher education. This align with Crompton and Burke [22] who in their study considered adoption of AI in higher education in seven continents affirms that only three (2%) studies came from Africa within the year 2016 to 2022 when compared with other continents. Similarly, the uneven distribution of reports across the African countries represented in the review, with South Africa contributing the highest proportion at 33.3% and other countries like Lesotho, Egypt, Zambia, Nigeria, and Ghana each contributing only a single report (8.3%), may be attributed to variations in research capacity, funding, and institutional support for AI integration studies within the higher education systems of these nations. One primary concern is that STEM academics in African higher education could discourage its usage instead of engaging in AI literacy activities among the African higher education community.

Table 2 Overview of studies included in the review
Full size table
Fig. 3
figure 3

Distribution of reports on the usage of AI in African higher education

Full size image

Figure 4 shows the distribution of published reports per year. In the year 2015 to 2021, there were no reports to show research activities on the usage of AI by academics and undergraduates in African Higher Education for STEM education, however, 8.3% of the reports considered was published in the year 2022. This indicates that before 2022, AI adoption in African higher education was not popular or embraced. In 2023, there was an increase in the number of reports published by 25.0%, as four reports which represented 33.3% were identified. Similarly, within the first half 2024, the rate also increased by 25.0% at the time of this review, seven reports which represent 58.3% were identified. This could indicate a progression in the adoption of AI by STEM undergraduates and academics in African higher education. Hence, there is a need for guidance and monitoring on its ethical usage.

Fig. 4
figure 4

Distribution of publications per year

Full size image

The variations observed in Fig. 5 regarding the focus of the published reports on different STEM populations (8% on STEM academics, 67% on STEM undergraduates, and 25% on both STEM academics and undergraduates) may be indicative of the research priorities and funding allocations within the field of AI integration in African higher education. The higher percentage of reports focused on STEM undergraduates (67%) could suggest a stronger emphasis on integrating AI technologies into undergraduate curricula and assessing their impact on student learning and engagement. Conversely, the lower percentage of reports focused on STEM academics (8%) may reflect a relative lack of research on the professional development and adoption of AI tools by faculty members, which could be an important area for future investigation. This also raises concerns about the limited number of studies examining the purposeful use of AI by STEM academics and undergraduates in Africa. Research that focuses explicitly on the purposeful use of AI for STEM education in African higher education institutions is urgently needed.

Fig. 5
figure 5

Distribution of reports published on academic and undergraduate AI usage

Full size image

Figure 6 shows the distribution of goals for using AI tools among STEM academics and undergraduates in African higher education. 41.7% of the reports showed that undergraduates utilize AI tools for learning, while 16.7% showed that academics use AI tools for teaching. 16.7% of the reports considered confirmed that STEM undergraduates used AI tools for interactive academic engagement, while only 8.3% showed that academics used AI tools for academic engagement. Also, 16.7% reports indicate that undergraduate students and academics utilize AI tools primarily for information search. In addition, 25.0% indicated that undergraduates’ goal of using AI tools is to generate academic content, while 16.7% show that academics also use AI tools for generating content. Moreso, 16.7% of the reports showed that academics and undergraduates use AI tools for research. Only 8.3% report indicated that undergraduates utilize AI tools to complete assignments, while 8.3% also reported that academics use AI tools to complete tasks. Additionally, 25.0% of the reports indicated that undergraduates utilize AI tools for academic purposes, while only 8.3% disclosed that academics use these tools for academic purposes. The disparities observed suggest that there may be differences in the priorities, needs, and adoption patterns between these two groups. The higher percentage of reports showing that undergraduates utilize AI tools for learning (41.7%) compared to academics using them for teaching (16.7%) could indicate a stronger focus on integrating AI-powered technologies into the student learning experience, potentially to enhance engagement, content generation, and information search. Conversely, the relatively lower percentage of reports on academics using AI tools for academic engagement (8.3%) and task completion (8.3%) may reflect a slower pace of academics adoption or a lack of targeted professional development and support for incorporating these technologies into their teaching and research practices.

Fig. 6
figure 6

Distribution of academics’ and undergraduates’ goals for using AI tools

Full size image

Figure 7 shows the distribution of forms of AI tools used by STEM academics and undergraduates. It showed that 83.3% of the reports indicated that ChatGPTs/chatbot/generative AIs were commonly used by undergraduates, while only 25.0% revealed that academics utilize ChatGPTs. The significantly higher percentage of reports indicating the use of ChatGPTs/chatbots/generative AI by undergraduates (83.3%) compared to academics (25.0%) could suggest that these emerging AI technologies are more readily integrated into student learning activities, potentially to enhance content generation, information search, and interactive engagement. The lower adoption of these tools among academics may be attributed to factors such as concerns about academic integrity, unfamiliarity with the technologies, or a lack of institutional support and guidelines for their appropriate use in teaching and research.Also, 16.7% indicated that undergraduates use personalized learning platforms, while 8.3% showed that academics use such platforms. 25.0% showed that undergraduates utilize grammar checkers, while 16.7% indicated that academics also utilize the tool. 41.7% indicated that undergraduates utilize other AI tools (such as Google Cloud, Wordtune, Turnitin, MATLAB, and Quillbot), while 33.3% indicated that academics utilize these other tools. Additionally, the disparities in the reported use of personalized learning platforms and grammar checkers between undergraduates and academics may indicate differing needs and preferences for AI-driven tools that support learning, writing, and productivity.

Fig. 7
figure 7

Distribution of forms of AI tools used by academics and undergraduates

Full size image

12 Discussion

12.1 Goals for utilizing AI tools by academics in STEM education

STEM academics have three obligations within higher education communities: engaging in teaching, research, and community service in line with science, technology, engineering, and mathematics. They are the ones responsible for the teaching and mentoring of undergraduates. This teaching involves three stages: planning/preparation, presentation, and evaluation. The planning stage involved the preparation of lecture notes, which required consultation on some materials and critical thinking toward providing meaningful content. Likewise, during the evaluation stage, academics are responsible for formulating several questions for students to answer. These factors combined can be demanding and stressful for academics, especially considering the presence of deadlines. As a result, they may be inclined to seek more efficient ways to complete the various tasks at hand. Hence, academics consider the adoption of AI for various purposes. The goals for the adoption of AI by academics include student learning and engagement, research and administrative processes [65], information searches, self-learning, content generation, paraphrasing academic content [87], academic content generation and grammar checks [64], and teaching and research [53]. This indicates that academics mainly utilize AI tools for teaching, content generation, and research. This could indicate that most academics who use AI tools use them for research and teaching, which are their core responsibilities in higher education institutions. For research, they tend to use tools for generating and paraphrasing content, while for teaching purposes, they use AI tools to develop the content to be delivered during classroom instruction.

12.2 Goals for undergraduates’ adoption of AI tools in STEM education

Undergraduates in STEM are involved in several activities within their academic lives, including attending lectures, taking notes, completing assignments, carrying out practical work, participating in group work mostly in the laboratory, preparing and writing exams, attending and presenting at seminars, and conducting final year projects, among others. These tasks can be highly demanding for STEM undergraduates, prompting them to constantly seek more efficient ways to accomplish them. The use of AI by undergraduates has become so popular that many people cannot perform academic tasks without engaging in AI [50]. STEM undergraduates utilize AI for many reasons, as described in the review. AI tools are used for learning and engagement in academic activities [12, 16, 29], information searches; self-learning; content generation; assignment; paraphrasing academic content [87], academic purposes [9], academic content generation and grammar checks [64], learning and research [53], enhancing academic writing [15], planning of lessons and classroom activities [84], completing assignments [1], and academic purposes [8]. Figure 6 shows that the main purpose of undergraduate use of AI tools is learning, closely followed by generating academic content and other academic purposes. These academic purposes could involve completing assignments, conducting information searches, and researching. This suggests a move away from more conventional learning methods, research, content generation, and academic writing and toward greater dependence on AI technologies for various academic activities. It also prompts worries about the possible effects on creativity, critical thinking, and ethical issues in academic work, so careful integration of AI technologies into educational frameworks is necessary to guarantee their appropriate and efficient usage.

12.3 The AI tools commonly used by STEM academics and undergraduates

There are several AI tools used by STEM academics and undergraduates in the literature, including virtual reality and personalized learning platforms [12, 65], ChatGPT/Chatbots/Generative AI [1, 8, 9, 15, 16, 29, 53, 64, 84, 87], grammar checker [15, 53, 64], Quillbot [15], Turnitin, MATLAB, Google Cloud, and WordTune [53]. Among all the AI tools identified in the literature, the ChatGPT emerges as a frequently utilized tool by undergraduates, as evidenced by most of the reports considered. This highlights the popularity and prominence of ChatGPT in supporting various academic activities and engaging students in AI-based interactive conversational experiences. The findings also reveal the wide variety of AI tools used by STEM academics and undergraduates, indicating the growing adoption of technology for enhanced learning and educational processes. Together with specialized software such as Quillbot, Turnitin, MATLAB, Google Cloud, and WordTune, these AI technologies commonly used by academics and undergraduates include chatbots, generative AI systems, virtual reality platforms, and personalized learning platforms. This demonstrates how AI technologies are increasingly incorporated into STEM education, giving STEM academics and undergraduates access to various tools to help with learning, research, generating academic content, and writing projects.

12.4 Benefits and challenges of adopting AI tools in STEM education in African higher education institutions

The integration of AI technology in STEM education within African higher education institutions comes with both excitements and worries. The excitement comes from the benefits it brings and the worries based on the challenges it poses. Among the benefit of integrating AI tools in higher education as identified in the study are individualized learning experiences and improved student learning and engagement [12, 29, 65], enhanced academic performance and advancement [12, 29], improved research and administrative processes [65], and increased awareness and usage of generative AI for information retrieval, self-learning, and content generation [87]. This indicates that the acceptance of AI-powered tools may transform the learning experiences of students, therefore bringing in a period of personalized, interesting, and adaptable learning. Personalised learning systems, virtual reality, and chatbots could open fresh doors of academic success where students are free to flourish and shine at their own pace. Studies showing improvement in student performance and advancement illustrate the transforming power of these AI-driven tools [12, 29].

The challenges identified from the study include lack of institutional support and infrastructure [53], Limited training opportunities and knowledge gaps [53, 64], Concerns about ethical implications and the need for regulatory frameworks [64], Varying levels of adoption and usage among different disciplines and socioeconomic backgrounds [87, 65]. Basically, the fair distribution of these revolutionary technologies is hampered by the absence of institutional support and strong IT infrastructure within many African institutions. High-speed internet, technical knowledge, and financing are all absent without which the digital gap deepens and certain institutions and students devoid of the advantages of AI-enhanced learning [53]. Nonetheless, the challenges impeding the effective use of AI in African higher education need to be resolved by strategic foresight and collaborative efforts.

Furthermore, the lack of training possibilities and knowledge gaps among academics and students aggravate the difficulty as the successful integration of AI tools depends on thorough skill development and capacity-building activities [53, 64]. Add to these pragmatic challenges the ethical questions about the use of artificial intelligence in the classroom. Critics contend that the hazards of algorithmic prejudice, data privacy violations, and possible abuse call for the immediate creation of laws responsive to culture and regulatory systems. Ignoring these important ethical issues can erode confidence, discredit AI-driven activities, and jeopardise the long-term viability of these transforming technologies in the African higher education [64]. Ultimately, this is further complicated by the unequal acceptance and application patterns of AI across several fields and group of persons. User friendly policies and support systems are crucial to guarantee that the advantages of AI-enhanced STEM education are fairly shared and reachable to everyone [87, 65]. Essentially, AI has great potential for African higher education; yet, effective integration of this tool calls for a multifarious and collaborative strategy. African universities may use the transforming potential of AI to open new horizons of academic excellence, research innovation, and inclusive learning experiences for their students by tackling the infrastructure, capacity-building, ethical, and equity-related issues.

13 Limitations and future research

Firstly, the sample size of 12 reports included in the analysis, although providing valuable insights, may not be fully representative of the broader landscape of AI integration in African higher education. The uneven distribution of reports across different countries and the potential underrepresentation of publications in non-English languages could have introduced biases in the data. Future research would benefit from a more comprehensive and systematic search strategy to capture a larger and more diverse set of publications, including those in local African languages, to gain a more holistic understanding of the topic. Additionally, the review focused primarily on the reported usage and goals of AI tools among STEM academics and undergraduates but did not delve deeply into the specific implementation approaches, pedagogical strategies, or learning outcomes associated with these technologies. Further research maybe needed to explore the practical applications of AI in teaching and learning, the barriers and enablers to effective integration, and the impact on student engagement, academic performance, and learning outcomes. Another limitation is the reliance on secondary data from published reports, which may not capture the nuances and contextual factors influencing AI integration in different African higher education institutions. Future studies could employ primary data collection methods, such as surveys, interviews, or case studies, to gain a more in-depth understanding of the experiences, perceptions, and challenges faced by both faculty and students in adopting and utilizing AI tools. Finally, the current review focused on the use of AI tools in STEM disciplines, but there is a need to investigate the integration of these technologies across a broader range of academic fields and disciplines within African higher education. Expanding the research to include the humanities, social sciences, and other areas could provide a more comprehensive picture of the opportunities and challenges associated with AI integration in the diverse educational landscape of the continent.

14 Conclusion and recommendations

AI tools have changed how things are done virtually in all human endeavours, with the educational sector being included in this transformation. In African higher education, academics and undergraduates utilize AI tools for various purposes. According to the review, academics’ strategic goal for adopting AI is to enhance the instructional process and research activities. AI tools are utilized primarily by STEM undergraduates for learning purposes, content generation, and various academic activities, such as assignment completion, information search, and research. ChatGPT has emerged as a commonly used tool among STEM undergraduates because of its effectiveness in facilitating interactive and conversational experiences. Based on the findings of this review, it is recommended that educational institutions in Africa acknowledge and embrace the integration of AI tools in higher education, given their potential to enhance learning outcomes and academic engagement. Students should be provided with guidelines and resources to ensure their responsible and ethical use of AI tools, emphasizing the importance of critical thinking, originality, and academic integrity. Academics should receive adequate training and support to effectively incorporate AI tools into their instructional practices, as this training will enable them to leverage the benefits of these tools while maintaining a balance with traditional pedagogical approaches. In addition, there should be an urgent focus on AI literacy across African higher education to advocate for the effective adoption of AI tools within the system. Continuous research and evaluation of the adoption and impact of AI on STEM education in Africa should be encouraged and conducted.