Project – The Impact of Artificial Intelligence on Academic Performance and Learning Efficiency among University Students: A Case Study of two Selected Higher Institutions in Lagos

CHAPTER ONE

INTRODUCTION

1.1 Background of the Study

Education has traditionally been regarded as one of the most important instruments for individual development, social transformation and national development. At the university level, education is expected not only to transmit knowledge but also to develop students’ ability to think critically, solve problems, communicate effectively, conduct research and apply knowledge to real-life situations. Consequently, universities continually adopt new technologies and pedagogical approaches that can improve teaching, learning and academic achievement.

The rapid development of digital technology has significantly transformed the higher education environment. Information and communication technologies have changed how students access information, communicate with lecturers, collaborate with colleagues and complete academic tasks. More recently, Artificial Intelligence (AI) has emerged as one of the most influential technologies affecting higher education. AI refers broadly to computational systems capable of performing tasks that are normally associated with human intelligence, including learning, reasoning, language processing, problem-solving, prediction and decision-making.

The application of AI in education is not entirely new. Before the recent emergence of generative AI, educational institutions had already used intelligent tutoring systems, automated assessment systems, adaptive learning platforms, recommendation systems, plagiarism-detection software and learning analytics. Crompton and Burke (2023), in a systematic review of AI in higher education, demonstrate that AI applications had already become relevant to areas such as assessment, prediction, personalisation, intelligent tutoring and administrative support.

However, the emergence of generative artificial intelligence (GenAI) has substantially increased the visibility and accessibility of AI among university students. Generative AI systems are capable of producing text, explanations, summaries, computer code, images and other forms of content in response to user prompts. Among the most prominent examples is ChatGPT, introduced publicly in November 2022. Lo (2023) notes that ChatGPT rapidly attracted attention in education because of its ability to generate coherent, human-like responses to users’ questions and instructions.

The emergence of ChatGPT and related AI tools has created significant opportunities for students. A student can use an AI system to obtain explanations of difficult concepts, generate examples, summarise information, receive assistance with writing, brainstorm research topics, obtain feedback on drafts, translate text and practise questions. Consequently, AI has the potential to function as an additional learning resource available to students beyond the conventional classroom environment.

One of the major potential benefits of AI in education is personalised learning. Traditional classroom instruction generally requires a lecturer to teach a group of students at the same time, even though students differ considerably in their prior knowledge, learning speed and learning preferences. AI tools can provide explanations at different levels of complexity and respond to individual questions. This creates the possibility of students receiving learning support according to their immediate needs.

AI may also improve learning efficiency. Learning efficiency can be understood as the ability of students to achieve intended learning outcomes with effective use of available time, cognitive resources and learning materials. Students may use AI to locate relevant information quickly, simplify difficult concepts, generate practice questions, identify errors in their work and receive immediate explanations. Such functions may reduce the amount of time students spend on routine academic tasks and allow them to devote more attention to higher-level learning activities.

Empirical research has increasingly examined these possibilities. Deng et al. (2024), in a systematic review and meta-analysis of experimental studies, found that ChatGPT interventions generally improved academic performance, affective-motivational outcomes and higher-order-thinking propensities while reducing mental effort. However, the authors also cautioned that methodological limitations in existing studies require careful interpretation of the findings.

More recent evidence is also encouraging. A 2026 meta-analysis of 66 studies and 129 effect sizes found a substantial positive overall effect of ChatGPT applications on undergraduate students’ learning outcomes, although the researchers examined studies published between 2023 and 2025 and noted the diversity of contexts and applications involved. This emerging evidence suggests that AI can potentially improve learning outcomes when used appropriately.

AI can support students’ academic performance in several ways. First, it can provide rapid access to explanations of academic concepts. A student who encounters difficulty understanding a concept during personal study can ask an AI tool to explain the subject using simpler language or practical examples. Second, AI can assist students with brainstorming and organising ideas. Third, it can provide feedback on writing and identify grammatical or structural problems. Fourth, it can generate practice exercises that students can use for revision.

AI can also support research activities. University students frequently encounter challenges in identifying research topics, understanding complex academic literature, structuring assignments and improving academic writing. AI tools can assist with brainstorming, summarisation and preliminary organisation of information. However, students must independently verify AI-generated information because generative AI can produce inaccurate or fabricated content.

The usefulness of AI therefore depends partly on how students use it. AI can function as a learning assistant when students use it to clarify concepts, generate questions, obtain feedback and deepen understanding. On the other hand, it can become a substitute for learning when students simply request complete answers and submit AI-generated content without understanding or evaluating it.

This distinction is important because academic performance is not merely about completing assignments. University education seeks to develop knowledge, reasoning, independent thinking, problem-solving and communication skills. If students rely excessively on AI to perform intellectual tasks that they are expected to learn themselves, the technology could potentially weaken some of the educational outcomes that universities seek to achieve.

The emergence of AI has therefore created a dual effect in higher education. On one hand, AI can improve accessibility, speed, personalisation and academic support. On the other hand, excessive or inappropriate use can create problems involving overdependence, reduced independent learning, misinformation, academic dishonesty and weakening of critical thinking.

This duality is increasingly recognised in the academic literature. Kasneci et al. (2023) argue that large language models such as ChatGPT present significant opportunities for education but also create important challenges concerning accuracy, bias, academic integrity and the changing nature of teaching and learning. Similarly, UNESCO (2023) emphasises that generative AI should be integrated into education through a human-centred approach that considers ethical, pedagogical, privacy and equity concerns.

The issue of academic integrity is particularly important. Students can use generative AI to generate essays, answer examination-style questions, write computer code and complete other assignments. If students present AI-generated work as their own without appropriate disclosure or understanding, the distinction between legitimate technological assistance and academic misconduct becomes increasingly difficult to establish.

UNESCO (2023) therefore recommends that educational institutions develop appropriate policies, build human capacity and establish responsible approaches to generative AI in education and research. The rapid development of AI means that universities cannot simply ignore its existence; instead, they need to understand how students are using these technologies and establish educational practices that encourage beneficial use while limiting harmful practices.

Another major concern is critical thinking. Critical thinking requires students to analyse information, evaluate evidence, identify assumptions and reach reasoned conclusions. If students consistently depend on AI to provide ready-made answers, they may have fewer opportunities to practise these skills. However, the opposite may also occur: students can use AI-generated responses as starting points and critically evaluate them, thereby creating additional opportunities for analysis and discussion.

Recent empirical research supports the importance of this distinction. A study of university students by Zhang et al. (2024) examined both the causes and consequences of generative AI use and highlighted the need to understand whether AI use produces beneficial or harmful outcomes for students rather than assuming that its effects are automatically positive or negative.

Another important consideration is student acceptance and actual use of AI. Students are more likely to use AI when they perceive it as useful, accessible and compatible with their academic tasks. Strzelecki (2023) examined ChatGPT acceptance among higher-education students and demonstrated that technology-adoption factors are relevant to understanding students’ willingness to use the tool. Similarly, a study involving 1,117 higher-education students found that facilitating conditions and behavioural intention were important determinants of ChatGPT use for learning. (Rachmawati et al., 2023).

The increasing use of AI also raises questions concerning learning efficiency. University students frequently experience pressure from multiple academic requirements, including lectures, assignments, examinations, projects and research activities. AI may reduce the time required to perform some routine activities. For example, students may use AI to summarise lengthy materials, generate outlines, explain unfamiliar terms or provide practice questions.

Nevertheless, speed should not automatically be equated with learning efficiency. A student who completes an assignment quickly through AI assistance without acquiring the underlying knowledge may have saved time but failed to achieve the intended learning outcome. Therefore, learning efficiency should involve both effective use of time and meaningful acquisition of knowledge and skills.

AI’s potential to improve academic performance may also depend on the quality of its outputs. Generative AI systems do not always provide accurate information. They may produce fabricated citations, outdated information or plausible but incorrect explanations. This creates a particular problem for university students who may not possess sufficient subject knowledge to identify errors.

Consequently, AI literacy has become increasingly important. Students need to know not only how to operate AI tools but also how to formulate effective prompts, verify information, recognise limitations, protect personal data and use AI ethically. UNESCO (2023) emphasises the importance of human capacity development and responsible use in integrating generative AI into education.

The issue is especially significant in developing educational contexts such as Nigeria. Nigerian universities are operating in an environment characterised by growing digitalisation, increasing internet access, mobile technology use and expanding student exposure to international digital platforms. Nigerian students can access AI tools from smartphones and computers, making AI increasingly difficult to separate from everyday academic life.

Recent Nigerian research has begun to examine the impact of AI on university students. A 2026 study at the University of Nigeria, Nsukka, examined students’ awareness, knowledge and use of AI technologies and identified concerns about AI dependency and its possible effects on reading culture.

Similarly, a recent study of students at Adeyemi Federal University of Education, Ondo, Nigeria, reported high awareness and substantial use of ChatGPT among students, with reported applications including research, essay writing and grammar checking. These findings indicate that AI is already becoming part of Nigerian students’ academic activities.

The Lagos context is particularly relevant. Lagos is one of Nigeria’s most important centres of higher education, technology, business and innovation. Students in Lagos have access to numerous digital services and are likely to encounter AI applications through academic, professional and personal activities. Consequently, examining AI use among students in selected higher institutions in Lagos can provide useful evidence about how this technology is affecting university learning.

The present study focuses on two selected higher institutions in Lagos. The purpose is not to assume that AI has only positive or negative consequences but to investigate its actual perceived impact on students’ academic performance and learning efficiency.

The concept of academic performance refers to the level of achievement demonstrated by students in their academic activities. It may be reflected through examination results, assignment performance, test scores, understanding of course materials and ability to complete academic tasks successfully. AI may influence academic performance by providing students with additional learning support, but its effect may also depend on the quality and manner of use.

Learning efficiency, on the other hand, relates to how effectively students use time and learning resources to acquire knowledge and skills. AI may improve efficiency by providing immediate responses, simplifying complex concepts and reducing the time spent on routine academic activities. However, efficiency can be undermined when students become overly dependent on AI or accept its responses without critical evaluation.

The relationship between AI and academic performance is therefore complex. It cannot be assumed that students who use AI frequently will automatically achieve higher grades. Frequency of use, purpose of use, level of AI literacy, type of academic task, students’ prior knowledge and the extent of independent engagement may all influence the outcome.

A 2024 study of university students found that ChatGPT usage was associated with both potential benefits and harmful consequences, illustrating the importance of examining the conditions under which generative AI is used. Similarly, research on ChatGPT in higher education has identified opportunities for personalised learning and academic support alongside concerns about accuracy, academic integrity and human interaction. (Bhullar, Joshi, & Chugh, 2024).

There is also increasing evidence that AI can support higher-order learning when it is integrated appropriately into educational activities. Deng et al. (2024) found positive effects on higher-order-thinking propensities in experimental studies, suggesting that AI does not necessarily undermine higher-level learning when used within appropriate educational interventions.

At the same time, the technology may change the role of the lecturer. Rather than serving only as a source of information, lecturers may increasingly function as facilitators, mentors and evaluators who help students interpret, verify and apply AI-generated information. This means that AI adoption requires changes not only in student behaviour but also in teaching strategies and assessment practices.

Universities are consequently confronted with the challenge of determining whether AI should be prohibited, restricted or integrated into teaching and learning. Evidence from university policy research indicates considerable variation in institutional responses. Xiao, Chen and Bao (2023), for example, found substantial differences in how leading universities responded to ChatGPT, with institutions adopting approaches ranging from bans to more permissive integration.

The Nigerian higher education environment similarly requires evidence-based approaches. Blanket prohibition may be difficult to enforce because AI tools are widely accessible, while unrestricted use may expose students and institutions to academic-integrity and learning-quality problems. Understanding students’ actual experiences is therefore essential.

Another issue is equity of access. Not all students necessarily have equal access to reliable internet connections, smartphones, computers or paid AI services. If AI becomes an important learning resource, differences in access could create inequalities between students. UNESCO (2023) specifically identifies inclusion and equity among the issues that should be considered in the responsible adoption of generative AI in education.

Furthermore, students’ digital literacy may influence the outcomes of AI use. A digitally literate student may critically evaluate AI-generated information, compare it with textbooks and scholarly sources, refine prompts and use AI as a supplementary learning tool. A less digitally literate student may simply copy AI-generated answers without verification. Thus, the same technology can produce different educational outcomes for different users.

The study is also justified by the rapidly changing nature of AI technology. Unlike many established educational technologies, generative AI is developing extremely quickly. New models and applications are continuously being introduced, making findings from earlier periods potentially less applicable to current students. This creates a need for continuing empirical research in specific institutional and geographical contexts.

Recent meta-analytic evidence provides further justification. Deng et al. (2024) concluded that ChatGPT can enhance academic performance and some learning-related outcomes, but they also recommended further research using stronger designs, longer-term measures and more complex assessments. The 2026 meta-analysis similarly found positive learning effects but drew evidence from a heterogeneous body of studies.

Therefore, the central issue is no longer simply whether artificial intelligence exists in higher education. The more important question is how students use AI, what educational benefits they obtain from it, and whether its use actually improves academic performance and learning efficiency.

This study is consequently designed to investigate the impact of artificial intelligence on academic performance and learning efficiency among university students in two selected higher institutions in Lagos. The study will provide empirical evidence that may assist students, lecturers, university administrators and education policymakers in developing responsible and effective approaches to AI-supported learning.

1.2 Statement of the Problem

The rapid emergence of artificial intelligence has created a significant transformation in higher education. University students can now use AI tools to generate explanations, summarise materials, write and edit academic content, solve problems, generate ideas and obtain immediate feedback. While these capabilities have the potential to improve students’ learning experiences, they have simultaneously created uncertainty concerning their effects on academic performance, independent learning and learning efficiency.

The first major problem is the uncertainty regarding whether AI actually improves students’ academic performance. Although students may perceive AI as useful, perceived usefulness does not necessarily translate into improved academic achievement. Some students may use AI to understand difficult concepts and improve their work, while others may use it primarily to generate answers that they submit without adequate engagement. Therefore, empirical investigation is necessary to determine the extent to which AI use is associated with academic performance.

The second problem is overdependence on artificial intelligence. The convenience of receiving instant answers may encourage some students to rely on AI rather than engage directly with textbooks, scholarly articles, lectures and independent problem-solving. Such dependence may reduce opportunities for students to develop the intellectual skills that university education is expected to cultivate.

The third problem is the potential effect of AI on critical thinking and independent learning. University students are expected to evaluate information, construct arguments and solve problems independently. If students routinely outsource these activities to AI systems, there is concern that their intellectual engagement may be weakened. However, AI may also be used to stimulate critical thinking when students evaluate, challenge and improve AI-generated responses. The actual outcome therefore depends on how the technology is used.

The fourth problem is academic dishonesty and integrity. Generative AI can produce essays, reports, solutions and other academic materials within seconds. Students may therefore use AI to complete assignments without acknowledging its contribution or demonstrating their own understanding. UNESCO (2023) identifies academic integrity, ethical use and responsible governance as important issues in the adoption of generative AI in education.

The fifth problem is the reliability and accuracy of AI-generated information. Generative AI systems can produce convincing but incorrect information. Students who accept AI-generated responses without verification may unknowingly incorporate errors into assignments, research projects and examinations. This could negatively affect both academic performance and the quality of students’ knowledge.

The sixth problem concerns learning efficiency. AI has the potential to reduce the amount of time required to search for information, organise ideas and obtain explanations. However, faster completion of academic tasks does not necessarily mean that students have learned more effectively. A student may complete an assignment rapidly through AI assistance but acquire little knowledge from the process. Consequently, it is necessary to examine whether AI improves genuine learning efficiency rather than merely reducing task-completion time.

The seventh problem is the lack of adequate institutional clarity concerning AI use. Universities are still developing policies concerning acceptable and unacceptable applications of generative AI. Research on university policies demonstrates considerable variation in institutional responses, with some institutions restricting AI while others encourage controlled educational use. (Xiao et al., 2023). This uncertainty can leave students and lecturers unclear about the boundaries of acceptable AI-assisted academic work.

The eighth problem is unequal levels of AI literacy among students. Students differ in their ability to use technology effectively and critically. Some students may understand how to verify AI-generated information and use AI to support learning, whereas others may simply copy generated responses. These differences can influence the educational outcomes associated with AI.

The ninth problem is limited context-specific evidence from Nigerian higher institutions, particularly concerning students in Lagos. Although international research on AI and higher education is expanding rapidly, educational environments differ across countries because of differences in institutional policies, technological infrastructure, student characteristics, teaching practices and access to digital resources.

Emerging Nigerian studies have begun to provide useful evidence. Research at Adeyemi Federal University of Education, Ondo, reported high awareness and use of ChatGPT among students, particularly for research, essay writing and grammar checking. A recent study at the University of Nigeria, Nsukka, has also examined awareness, knowledge and use of AI technologies among students while raising concerns about AI dependency and reading culture. These studies indicate the need for additional research across different Nigerian higher-education contexts.

The tenth problem is that existing findings are not entirely consistent concerning the educational consequences of AI. Some research demonstrates positive effects on academic performance and learning outcomes, while other literature highlights risks involving overreliance, academic integrity and reduced independent engagement. Deng et al. (2024) found generally positive effects in experimental studies but also emphasised methodological limitations and the need for stronger long-term research.

This situation creates an important research gap. There is a need to move beyond general discussions about whether AI is “good” or “bad” for education and instead examine how AI is actually affecting students within specific educational environments.

The problem is therefore that artificial intelligence is becoming increasingly integrated into students’ academic activities, yet there is insufficient context-specific empirical evidence concerning its impact on students’ academic performance and learning efficiency in selected higher institutions in Lagos.

The present study consequently seeks to examine whether AI contributes positively to students’ academic performance and learning efficiency, the ways in which students use AI for academic purposes, and the challenges associated with its use.

1.3 Purpose of the Study

The general purpose of this study is to examine the impact of artificial intelligence on academic performance and learning efficiency among university students in two selected higher institutions in Lagos.

Specifically, the study seeks to:

1. examine the level of awareness and use of artificial intelligence among students in the selected higher institutions;
2. identify the major ways students use artificial intelligence for academic purposes;
3. examine the impact of artificial intelligence on students’ academic performance;
4. determine the impact of artificial intelligence on students’ learning efficiency;

1.4 Research Questions

The following research questions will guide the study:

1. What is the level of awareness and use of artificial intelligence among students in the selected higher institutions?
2. What are the major ways students use artificial intelligence for academic purposes?
3. To what extent does artificial intelligence affect students’ academic performance?
4. To what extent does artificial intelligence affect students’ learning efficiency?

1.5 Research Hypothesis

The following null hypothesis will be tested at the 0.05 level of significance:

H₀: Artificial intelligence has no significant impact on the academic performance and learning efficiency of university students in the two selected higher institutions in Lagos.

1.6 Significance of the Study

Students

The study will help university students understand both the benefits and limitations of AI as an academic resource. It may encourage students to use AI to support learning rather than replace independent study, critical thinking and academic effort.

Lecturers

The findings will provide lecturers with information about how students are using AI and the perceived effects of such use on learning. This may help lecturers redesign assignments, teaching strategies and assessment methods in ways that encourage responsible AI use.

University Management

University administrators may use the findings to develop or improve institutional policies on artificial intelligence. Such policies can distinguish between appropriate AI-assisted learning and academic misconduct.

Educational Policymakers

The findings may provide useful evidence for policymakers concerned with digital transformation and technology integration in Nigerian higher education.

Curriculum Developers

The study may assist curriculum developers in considering AI literacy, digital literacy, critical thinking and responsible technology use as components of contemporary university education.

Researchers

The study will contribute to the growing literature on artificial intelligence and higher education, particularly within the Nigerian context. It may serve as a reference for researchers conducting future studies on AI, academic performance, digital learning and educational technology.

Higher Education Institutions

The findings may help institutions identify appropriate ways of integrating AI into teaching and learning while reducing risks associated with misuse, overdependence and academic dishonesty.

1.7 Scope of the Study

The study focuses on the impact of artificial intelligence on academic performance and learning efficiency among university students in two selected higher institutions in Lagos.

The independent variable is artificial intelligence use, particularly students’ use of generative AI applications such as ChatGPT and related AI-based educational tools.

The dependent variables are:

• academic performance, and
• learning efficiency.

AI use will be examined in relation to activities such as:

• understanding difficult concepts;
• research and information gathering;
• assignment preparation;
• academic writing;
• summarisation;
• problem-solving;
• revision and examination preparation;
• grammar and language improvement; and
• generation of practice questions.

Academic performance will be considered in terms of students’ perceived or reported improvement in:

• test and examination performance;
• assignment quality;
• understanding of course content;
• ability to solve academic problems; and
• achievement of learning objectives.

Learning efficiency will be examined in relation to:

• time saved during academic tasks;
• speed of accessing learning resources;
• ease of understanding difficult materials;
• ability to organise academic work;
• productivity during study; and
• effectiveness of learning activities.

Geographically, the study is limited to two selected higher institutions in Lagos State. The study does not seek to generalise automatically to every university or higher institution in Nigeria.

1.8 Operational Definition of Terms

Artificial Intelligence (AI): Computer-based technologies designed to perform tasks associated with human intelligence, including learning, reasoning, language processing, prediction and problem-solving.

Generative Artificial Intelligence (GenAI): AI systems capable of generating new content such as text, images, computer code, summaries and responses based on user instructions.

ChatGPT: A generative AI conversational system developed by OpenAI that can respond to natural-language prompts and generate different forms of text-based content.

Academic Performance: The level of achievement demonstrated by a student through academic activities such as examinations, tests, assignments, projects and other assessments.

Learning Efficiency: The extent to which a student can effectively acquire knowledge and skills while making productive use of available time, effort and learning resources.

AI Use: The frequency and manner in which students employ artificial intelligence tools for academic and learning activities.

AI-Assisted Learning: A learning process in which students use artificial intelligence as a supplementary resource to support understanding, research, practice, feedback or other educational activities.

AI Literacy: The ability to understand, use, evaluate and critically interact with artificial intelligence technologies.

Academic Integrity: The principles of honesty, responsibility, originality and ethical conduct expected in academic work.

Critical Thinking: The ability to analyse information, evaluate evidence, identify assumptions and reach logical and reasoned conclusions.

University Student: An individual enrolled in an undergraduate or other recognised programme of study in a higher-education institution.

Learning Outcome: The knowledge, skills, competencies or abilities that a student is expected to acquire as a result of an educational experience.

AI Dependency: Excessive reliance on artificial intelligence to perform academic or intellectual activities that students would ordinarily be expected to perform independently.

Responsible AI Use: The ethical and academically appropriate use of artificial intelligence in ways that support learning while maintaining academic integrity, independent thinking, privacy and accountability.

1.9 Organisation of the Study

The study is organised into five chapters.

Chapter One presents the introduction, background of the study, statement of the problem, purpose of the study, research questions, research hypothesis, significance of the study, scope of the study, operational definition of terms and organisation of the study.

Chapter Two will review relevant literature on artificial intelligence, generative AI, academic performance, learning efficiency and AI-assisted learning. It will include the conceptual review, theoretical framework and empirical review.

Chapter Three will present the methodology of the study, including the research design, population, sample size, sampling technique, research instrument, validity, reliability, method of data collection and method of data analysis.

Chapter Four will present and analyse the data collected from respondents. It will include the demographic characteristics of respondents, presentation and analysis of data, analysis of research questions and testing of the research hypothesis.

Chapter Five will present the summary of findings, conclusion and recommendations. It will also discuss the contribution of the study to knowledge and provide suggestions for further research.

Project – The Impact of Artificial Intelligence on Academic Performance and Learning Efficiency among University Students: A Case Study of two Selected Higher Institutions in Lagos