✨
Trusted Academic Publisher
Serving Researchers Since 2012

Generative AI in Higher Education: A Multi-College Study of AI Adoption, Academic Integrity and Perceived Employability among Undergraduate Students in Bengaluru

DOI : 10.17577/IJERTCONV14IS110157
Download Full-Text PDF Cite this Publication

Text Only Version

Generative AI in Higher Education: A Multi-College Study of AI Adoption, Academic Integrity and Perceived Employability among Undergraduate Students in Bengaluru

Dr. Latha Thimmaiah

Associate Professor

Hod, Dept. Commerce & Management Government first grade college, Malleshwaram, Bangalore

-12 Email ID: shreylatha@gmail.com

Mr. Shridhara S Assistant Professor Dept. of Commerce

Hoysala Degree College, Nelamangala, Bengaluru- 123 Email ID: Shridhara.ticm@gmail.com

1. Abstract – Generative artificial intelligence (Gen AI) is increasingly becoming part of the higher education experience, offering students new ways to understand concepts, prepare for examinations and complete academic tasks. However, its growing use also raises concerns regarding academic integrity, overdependence, information accuracy and responsible use. This study examines the adoption of generative AI, its perceived learning benefits, academic-integrity awareness and perceived career relevance among undergraduate students in Bengaluru. The study adopted a quantitative, cross-sectional survey design using a structured Google Forms questionnaire. A non-probability convenience sample of 153 responses was analysed using descriptive frequencies and percentages. The findings indicate that 87.6% of respondents reported previous academic use of generative AI. ChatGPT was the most frequently selected tool (85.6%), followed by Google Gemini (42.5%). A total of 88.2% agreed or strongly agreed that generative AI helps them understand difficult academic concepts, while 77.8% perceived potential benefits for career-related skills. At the same time, 77.8% expressed concern that excessive dependence on AI could weaken independent thinking, and 57.5% recognised academic-integrity concerns associated with presenting AI-generated work as their own without following applicable rules. Only 32.0% reported always or often verifying AI-generated information, while 63.4% answered No or Maybe when asked about formal AI guidance or training. The study highlights the need for clear institutional guidelines, practical AI-literacy programmes, information- verification skills and transparent disclosure practices. The findings are exploratory and reflect the perceptions of the surveyed students rather than measured learning or employment outcomes.

Keywords: Generative artificial intelligence, higher education, undergraduate students, academic integrity, AI literacy, perceived career relevance, Bengaluru.

  1. INTRODUCTION

    The rapid development of Generative Artificial Intelligence (Gen AI) has introduced new possibilities for teaching, learning and knowledge creation in higher education. Tools powered by large language models can generate text, explain concepts, summarise information, support brainstorming and assist users with a wide range of academic tasks. Their accessibility has made them increasingly relevant to undergraduate students across disciplines, including commerce, management and computer applications. However, the educational value of these tools depends not only on their availability but also on how students understand, evaluate and use their outputs.

    Gen AI can support students in understanding difficult concepts, preparing revision materials, generating ideas for assignments and exploring unfamiliar topics. Kasneci et al. (2023) discuss both the educational opportunities of large language models and the need to address their limitations, including inaccurate outputs, bias, overreliance and the importance of critical evaluation. Therefore, AI- assisted learning should be considered a complement to students own reasoning and engagement rather than an automatic substitute for independent learning.

    The adoption of Gen AI also raises questions concerning academic integrity. Students may use these tools to clarify concepts or improve their writing, but concerns arise when AI-generated content is submitted as original work in violation of institutional rules, when sources are not verified, or when students cannot explain the work they submit. Research involving university students has examined differing attitudes towards AI-assisted academic work and the challenges that these technologies create for established approaches to academic integrity (University of Liverpool student perspectives study, 2024; The rapid rise of generative AI and its implications for academic integrity, 2024). These issues highlight the importance of clear institutional expectations, transport disclosure where required and education about responsible AI use.

    Another important consideration is students perceptions of the relevance of Gen AI to their future careers. As digital technologies become part of many professional activities, students may view AI literacy as a useful capability for their transition form college to employment. Nevertheless, perceived usefulness should not be treated as proof that AI use directly improves employability. Students must also develop subject knowledge, communication, problem-solving, ethical judgement and the ability to verify information. UNESCOs AI competency framework for students emphasises a human-centred mindset, AI ethics, knowledge of AI techniques and applications, and the ability to understand and engage with AI systems responsibly (Miao & Shiohira, 2024).

    In India, undergraduate students in commerce, management and computer applications may encounter different academic tasks and career expectations. BCom students may use Gen AI to support the understanding of accounting, finance and business concepts; BBA students may use it for marketing, management and business communication; and BCA students may use it for programming explanations and technical problem-solving. These are potential applications rather than assumptions about the behaviour of every student. Empirical evidence is required to understand actual usage patterns and whether perceptions differ across programmes.

    Against this background, the present study examines undergraduate students use of Gen AI in higher education in Bengaluru. It focuses on the purposes and frequency of AI use, perceived learning benefits, perceived usefulness for future careers, awareness of academic integrity, practices for verifying AI-generated information and access to formal AI guidance. By examining these dimensions together, the study aims to contribute to a balanced understanding of Gen AI as both a learning resource and a technology requiring responsible use.

    The findings are intended to inform students, faculty members and higher education institutions about practical measures for AI literacy, responsible use academic integrity and career-oriented digital learning. Rather than treating Gen AI adoption as inherently beneficial or harmful, the study approaches it as an educational development whose value depends on informed, critical and ethical engagement.

  2. BACKGROUND AND LITERATURE REVIEW

    1. Gen AI as a learning resource

      Kasneci et al. (2023) discuss how large language models can support educational explanations, engagement and learning innovation. They also emphasise the need for human oversight, critical thinking and verification of AI-generated content.

    2. Student adoption and academic integrity

      A University of Liverpool study collected responses from 2,555 students and explored their attitudes towards different uses of generative AI. The findings illustrate why students may accept some forms of AI assistance while opposing AI-generated work being submitted as a complete essay.

    3. AI literacy and responsible use

      UNESCOs student competncy framework presents AI education as more than simply learning to operate a tool. It includes ethical awareness, understanding AI systems and developing the ability to use them responsibly.

    4. AI use in assessments

      A 2024 survey of 337 Australian University students examined the use of chatbots for assessments and students interpretations of academic integrity. It highlights the need for clear guidance and thoughtful assessment design.

    5. Students AI knowledge and limitations

      A 2026 study addresses the difference between what undergraduate students need to know about Gen AI and what they actually understand about its capabilities, limitations and societal implications. This is relevant to questions about students ability to recognise inaccurate or misleading AI outputs.

    6. Research gap

      Existing research has examined students adoption of generative AI, perceived learning benefits, academic integrity and AI literacy in different higher education contexts. However, the relevance of these findings may vary according to institutional settings, student populations and academic disciplines. There is a need for context-specific evidence examining these dimensions together among undergraduate students in Bengaluru, particularly across commerce, management and computer applications programmes.

      The present study addresses this need by investigating students reported patterns of Gen AI use, perceived usefulness for learning and future careers, information-verification practices, academic integrity awareness and exposure to formal AI guidance. It also exposure whether these responses vary across programmes and year of study. The study provides a local, exploratory contribution rather than claiming to represent all undergraduate students in Bengaluru or India.

  3. STATEMENT OF THE PROBLEM

    The increasing availability of generative AI tools presents undergraduate students with opportunities to support academic learning and develop familiarity with emerging digital technologies. However, the extent to which students use these tools, verify their outputs, understand academic integrity expectations and perceive their relevance to future employment may differ across educational programmes and individual experiences. Without adequate understanding of these dimensions, institutions may find it difficult to formulate effective guidance for responsible AI integration. The present study therefore investigates Gen AI adoption, perceived learning and career benefits, and academic integrity awareness among undergraduate students in Bengaluru.

  4. OBJECTIVES OF THE STUDY

    • To examine the adoption and frequency of generative AI use among undergraduate students.

    • To identify the major academic activities for which students use generative AI tools.

    • To assess students perceptions of the usefulness of generative AI for learning and future career preparation.

    • To examine students awareness of academic integrity and responsible AI use.

    • To investigate students practices for verifying AI-generated information and their exposure to formal AI guidance.

    • To explore differences in AI usage and perceptions across undergraduate programmes and years of study.

    • To suggest measures for promoting responsible, ethical and educationally meaningful integration of generative AI in higher education.

  5. RESEARCH METHODOLOGY

    1. Research design

      The study adopts a quantitative, cross-sectional research design to examine undergraduate students use and perceptions of generative AI in higher education. Data are collected through a structured online questionnaire.

    2. Study area and respondents

      The study focuses on undergraduate students in Bengaluru, Karnataka. The intended respondent groups include students pursuing Bachelor of Commerce (BCom), Bachelor of Business Administration (BBA) and Bachelor of Computer Applications (BCA), and other.

    3. Sampling technique and sample size

      A non-probability convenience sampling technique is used, as the questionnaire is circulated to students who are accessible and willing to participate. The final sample size is 153.

    4. Sources of data

      Primary data are collected through a structured Google Forms questionnaire. Secondary information is drawn from published academic research and institutional guidance on generative AI, higher education, academic integrity and AI literacy.

    5. Research instrument

      The questionnaire covers programme and year of study, experience and frequency of Gen AI use, tools used, academic applications, perceived learning benefits, perceived career relevance, verification practices, academic integrity awareness and exposure to formal AI training. It also includes an optional open-ended question on the benefits and challenges of Gen AI in higher education.

    6. Data analysis

      The responses will be exported to a spreadsheet and examined for completeness, consistency and potential duplicate entries where identifiable without collecting unnecessary personal information. Descriptive statistics, including frequencies and percentages, will be used to summarize responses. Cross-tabulations and charts will be prepared to compare relevant groups, such as programme and year of study.

    7. Ethical considerations

      Participation is voluntary. Respondents are informed about the purpose of the research, and responses are intended to be analysed and presented in aggregate form. Personal identifiers should not be collected unnecessarily. Any institutional permission required for the survey will be obtained, and the data will be used in accordance with applicable research and conference requirements.

  6. OVERALL RESULTS AND DISCUSSION

    Generative AI in Higher Education: A Multi-College Study of AI Adoption, Academic Integrity and Perceived Employability among Undergraduate Students in Bengaluru

    Sample: 153 responses

    1. Respondent profile

      Programme

      n

      %

      BCom

      121

      79.1%

      BCA

      16

      10.5%

      BBA

      12

      7.8%

      Other

      4

      2.6%

      Year of study

      n

      %

      Second year

      70

      45.8%

      Third year

      40

      26.1%

      First year

      32

      20.9%

      Other

      11

      7.2%

      College name (standardized

      from typed responses)

      n

      %

      Hoysala Degree College

      63

      41.2%

      SLN Degree College

      57

      37.3%

      Other institutions

      15

      9.8%

      Government First Grade College,

      Malleshwaram

      12

      7.8%

      Triveni College

      3

      2.0%

      VV Puram College

      3

      2.0%

      The study analysed 153 responses. BCom students accounted for 121 respondents (79.1%), followed by BCA students with 16 (10.5%) and BBA students with 12 (7.8%). Four respondents (2.6%) selected another programme or category. Second-year students formed the largest year-of-study group, with 70 respondents (45.8%), followed by third-year students with 40 (26.1%) and first- year students with 32 (20.9%). This profile indicates that the overall results primarily reflect the experiences of BCom respondents.

    2. Adoption, tools and academic uses

      Prior academic use of GenAI

      n

      %

      Yes

      134

      87.6%

      No

      11

      7.2%

      Not Sure

      8

      5.2%

      Figure 1. Previous academic use of generative AI.

      Tool selected (multiple

      responses allowed)

      Selections

      % of N

      ChatGPT

      131

      85.6%

      Google Gemini

      65

      42.5%

      Microsoft Copilot

      4

      2.6%

      Other AI tools

      15

      9.8%

      Selected no AI tool

      2

      1.3%

      Figure 2. AI tools selected by respondents. Percentages use N=153; options are not mutually exclusive.

      Academic activity (multiple

      responses allowed)

      Selections

      % of N

      Understanding difficult concepts

      83

      54.2%

      Exam preparation

      68

      44.4%

      Notes or summaries

      56

      36.6%

      Assignments/written work

      52

      34.0%

      Solving academic problems

      42

      27.5%

      Presentations

      39

      25.5%

      Project/research ideas

      0

      0.0%

      Figure 3. Main academic uses reported by students.

      A total of 134 respondents (87.6%) reported having previously used generative AI for academic purposes. ChatGPT was the most commonly selected tool, reported by 131 respondents (85.6%), followed by Google Gemini, selected by 65 respondents (42.5%). As multiple selections were permitted, these percentages are not mutually exclusive.

      The most frequently selected academic application was understanding difficult concepts, reported by 83 respondents (54.2%). Exam preparation was selected by 68 respondents (44.4%), followed by preparing notes or summaries by 56 (36.6%) and assignments or written work by 52 (34.0%). These results suggest that students in the sample primarily perceive generative AI as a learning and academic-preparation aid.

    3. PERCEIVED BENEFITS, ACADEMIC INTEGRITY AND AI LITERACY

      Statement

      Agree / strongly agree n

      (%)

      Neutral n (%)

      Disagree / strongly disagree

      n (%)

      Understanding difficult

      concepts

      135 (88.2%)

      15 (9.8%)

      3 (2.0%)

      Career-related skills

      119 (77.8%)

      27 (17.6%)

      7 (4.6%)

      Identify misleading output

      56 (36.6%)

      49 (32.0%)

      48 (31.4%)

      Integrity concern

      88 (57.5%)

      47 (30.7%)

      18 (11.8%)

      Know responsible use

      106 (69.3%)

      30 (19.6%)

      17 (11.1%)

      Concern about dependence

      119 (77.8%)

      26 (17.0%)

      8 (5.2%)

      Disclosure awareness

      62 (40.5%)

      50 (32.7%)

      41 (26.8%)

      A total of 135 respondents (88.2%) agreed or strongly agree that generative AI helps them understand difficult academic concepts. In addition, 119 respondents (77.8%) agree or strongly agreed that it can help develop career-related skills.

      These responses indicate a positive perception of the educational and career-related potential of GenAI. However, the findings represent students opinions and do not demonstrate actual improvements in academic performance, professional competence or employability.

      A total of 88 respondents (57.5%) agreed or strongly agree that submitting AI-generated work as ones own without following applicable rules raises academic -integrity concerns.

      Furthermore, 119 respondents (77.8%) expressed concern that excessive dependence on generative AI could weaken independent thinking.

      At the same time, 106 respondents (69.3%) agreed or strongly agreed that they know how to use AI without directly copying its output as their own. These findings suggest that students recognize several risks associated with Gen AI, although awareness and reported confidence do not necessarily guarantee consistent responsible behaviour.

      Only 62 respondents (40.5%) agree or strongly agreed that they understood the importance of disclosing AI use when required. This indicates scope for clearer communication about when disclosure is expected and how students should acknowledge AI assistance in accordance with institutional and assignment-specific rules.

      Regarding the ability to identify inaccurate or misleading AI-generated information, 56 respondents (36.6%) agreed or strongly agreed that they could identify such information, while 48 (31.4%) disagree or strongly disagreed. The remaining 49 respondents (32.0%) were neutral. These results indicate mixed levels of self-reported confidence and support the need for practical training in verification and critical evaluation.

      Figure 4. Selected indicators related to verification, integrity, dependence and formal training.

    4. Information Verification, disclosure and formal training

    Verification frequency

    n

    %

    Sometimes

    84

    54.9%

    Always

    27

    17.6%

    Often

    22

    14.4%

    Rarely

    16

    10.5%

    I have never used generative AI

    2

    1.3%

    Never

    2

    1.3%

    Formal guidance/training

    n

    %

    No

    59

    38.6%

    Yes

    56

    36.6%

    Maybe

    38

    24.8%

    Information verification emerged as an area requiring attention. Only 49 respondents 32.0%) reported that they always or often verify AI-generated information. The largest group, 84 respondents (54.9%), reported verifying information sometimes. This pattern suggests that checking AI outputs may not yet be a consistent practice for many respondents.

    When asked about formal AI guidance or training, 56 respondents (36.6%) answered Yes, 59 (38.6%) answered No and 38(24.8%) answered Maybe. Thus, 97 respondents (63.4%) either reported no formal guidance or were uncertain whether such guidance was avilable.

    The findings point to a practical opportunity for colleges to provide clear AI-use guidelines, faculty guidance, workshops and activities that tech students to verify generated information, protect privacy, acknowledge AI assistance and retain independent responsibility for submitted work.

    Overall, the results reveal a combination of substantial adoption, perceived learning value and concerns about verification, integrity and overdependence. The findings support a balanced educational approach that encourages beneficial use while developing responsible AI literacy.

  7. MAJOR FINDINGS

    • High reported Adoption: 87.6% of respondents reported previous academic use of generative AI.

    • Dominance of ChatGPT: ChatGPT was selected by 85.6% of respondents, making it the most commonly reported tool.

    • Learning Support: 88.2% perceived Gen AI as helpful for understanding difficult concepts.

    • Perceived career relevance: 77.8% believed Gen AI could help develop career-related skills.

    • Inconsistent verification: only 32 % reported always or often verifying AI-generated information.

    • Academic-integrity awareness: 57.5% recognized concerns about submitting AI-generated work as ones own without following applicable rules.

    • Concerns about overdependence: 77.8% were concerned that excessive reliance on AI could weaken independent thinking.

    • Mixed disclosure awareness: 40.5% agreed or strongly agreed that they understood the importance of disclosing AI use when required.

    • Need for institutional support: 63.4% answered No or Maybe regarding formal AI guidance or training.

  8. SUGGESTIONS AND IMPLICATIONS

    Based on the findings, the following measures are recommended for higher education institutions:

    • Develop clear institutional guidelines: Colleges should explain when Gen AI may be used, when it is restricted and when its use must be disclosed.

    • Conduct AI-literacy workshops: Students should receive practical instruction on prompt formulation, output evaluation, bias, privacy, citation and responsible use.

    • Promote information verification: students should be encouraged to cross-check AI-generated answers against textbooks, course materials, credible sources and faculty guidance.

    • Protect independent thinking: Assignments should encourage students to explain concepts in their own words, apply knowledge to real situations and demonstrate their reasoning.

    • Clarify academic-integrity expectations: Faculty members should provide examples of acceptable assistance inappropriate copying and transparent acknowledgement of AI-generated material.

    • Integrate career-relevant AI skills: Programme-specific demonstrations can show commerce, management and computer applications students how to use AI critically for relevant academic and professional tasks.

    • Strengthen faculty support: Teachers should receive guidance on responsible AI integration and the design of assessments that evaluate students understanding rather than only polished final answers.

    • Expand future research: Further studies should seek more balanced representation across programmes, years of study and colleges to examine whether the patterns differ across student groups.

  9. LIMITATIONS OF THE STUDY

    The study has several limitations. First, it used convenience sampling, and the sample was dominated by BCom students. Consequently, the findings should not be treated as representative of all undergraduate students in Bengaluru.

    Second, the data were self-reported and may be affected by recall and social-desirability bias.

    Third, the questionnaire measured students perceptions and reported practices rather than objectively assessed learning gains, academic performance or employment outcomes.

    Finally, the analysis was primarily descriptive and did not establish statistically significant differences between groups.

  10. CONCLUSION

    This study examined generative AI adoption, perceived learning benefits, academic-integrity awareness and perceived career relevance among 153 undergraduate student respondents in Bengaluru. The findings indicate that generative AI is widely used within the sample and is positively perceived as a tool for understanding difficult concepts and developing career-related skills. However, these perceived benefits coexist with concerns about excessive dependence, academic integrity and inconsistent verification of AI-generated information.

    The results also reveal a need for clearer institutional guidance and practical training in responsible AI use. Higher education institutions should therefore move beyond approaches based solely on permission or prohibition and develop learning environments that encourage critical evaluation, independent reasoning and transparent acknowledgement of AI assistance. Generative AI should complement, rather than replace, students own learning and judgement.

  11. REFERENCES

  1. Kasneci, E., Sessler, K., Kuchemann, S., Bannert, M., Demontieva, D., Fischer, F., et al. (2023). ChatGPT for good? On opportunities and challenges of large language models for education. https://doi.org/10.1016/j.lindif.2023.102274.

  2. Miao, F., & Holmes,W. (2023). Guidance for generative AI in education and research. UNESCO. https://doi.org/10.54675/EWZM9535.

  3. Gruenhagen, J.H., et al. (2024). The rapid rise of generative AI and its implications for academic integrity: Students perceptions and use of chatbots for assistance with assessments. Computers and Education: Artificial Intelligence, 7, 100273. https://doi.org/10.1016/j.caeai.2024.100273

  4. Bouteraa, M., Bin-Nashwan, S, A., AI-Daihani, M., Dirie, K. A., Benlahcene, A., Sadallah, M., et al. (2024). Understanding the diffusion of AI-generative (ChatGPT) in higher education: Does students integrity matter? Computers in Human Behavior Reports, 14, 100402. https://doi.org/10.1016/j.chbr.2024.100402

  5. van Dis, E. A. M., Bollen, J., Zuidema, W., van Rooji, R., & Bockting, C.L. (2023). ChatGPT: Five priorities for research. Nature, 614, 224-226. https://doi.org/10.1038/d41586-023-00288-7.

One thought on “Generative AI in Higher Education: A Multi-College Study of AI Adoption, Academic Integrity and Perceived Employability among Undergraduate Students in Bengaluru”

  1. vikram singh says:

    Nice multi-college study on generative AI adoption and academic integrity among students in Bengaluru. The link to perceived employability is thought provoking, and the policy recommendations are practical for institutions looking to adopt AI tools.

Comments are closed.