The qualitative study, conducted by researchers Mir Hasib of the University of Alabama and Md Shariful Islam of Khulna University, explored how university students perceive and use ChatGPT in their academic and personal lives. Through in-depth interviews with 20 students from Khulna University, the researchers found that the AI chatbot has rapidly become embedded in students’ daily routines, despite concerns over fabricated references, inaccurate information and ethical misuse.

The findings paint a picture of a generation that sees ChatGPT as a practical assistant rather than simply a search tool. Students described using it to complete coursework more efficiently, prepare presentation materials, generate ideas for research, polish writing, create social media content, write professional emails and cover letters, and even seek entertainment and lifestyle advice.

“Students are no longer using ChatGPT only as an academic tool,” the researchers wrote. “It has become a companion for learning, productivity and everyday decision-making.”

Assignments in minutes

One of the most common uses reported by participants was completing assignments and preparing classroom presentations.

Several students said ChatGPT helped them generate ideas, improve the quality of their writing and meet tight deadlines. Some asked the chatbot to rewrite assignments, while others used it to summarise lengthy materials into presentation notes.

One English student recalled using ChatGPT to write a short story in the style of Charlotte Brontë for a university assignment. After changing the names and setting to fit the Bangladeshi context, the student received one of the highest marks in the class.

Others used the chatbot to prepare speeches or compare complex legal frameworks in a fraction of the time they expected conventional research would require.

The study also found that many students now depend on ChatGPT instead of taking lecture notes. Rather than writing notes during class or borrowing notebooks from classmates before examinations, participants said they simply ask ChatGPT to generate concise summaries tailored to their needs.

Beyond Google and Grammarly

Students also reported replacing other digital tools with ChatGPT.

Instead of browsing multiple websites through traditional search engines, participants preferred receiving direct answers from ChatGPT. Some even said they had shifted from Google Search to Microsoft’s Bing because of its integrated AI assistant.

Others increasingly relied on ChatGPT instead of writing assistants such as Grammarly and QuillBot.

Interviewees cited the absence of strict word limits, flexible prompt-based editing and free access to advanced writing features as major advantages over competing platforms, many of which reserve key functions for paid subscribers.

Although students praised the chatbot’s speed and convenience, they also recognised its limitations.

Several participants said ChatGPT frequently generated references that did not exist, cited incorrect sources or produced generic research content that required extensive editing before it could be used.

One master’s student described receiving only one authentic citation after ChatGPT generated five references for a thesis introduction, while the remaining citations appeared to be fabricated.

Another participant said the chatbot repeatedly produced unnecessary sections of a research paper instead of answering specific methodological questions.

Job applications, emails and social media

Outside the classroom, students found ChatGPT equally useful.

Participants reported using it to write internship applications, prepare CVs, draft cover letters and compose dozens of emails to prospective supervisors abroad.

One student applying for graduate studies said ChatGPT significantly reduced the time required to write personalised emails to professors across North America and Europe.

Others used it to create captions for university sports pages, generate professional social media posts and rewrite CVs to better match job descriptions.

However, several interviewees noted that AI-generated CVs sometimes invented experiences or skills that were not part of their original documents, requiring careful manual correction before submission.

Some students also experimented with ChatGPT for leisure, asking it to recommend television series, generate jokes or suggest recipes based on available ingredients.

Academic integrity concerns

Despite appreciating ChatGPT’s usefulness, many participants openly acknowledged using it in ways that raise ethical questions.

Some admitted relying on the chatbot during assignments and open-book examinations, while others described deliberately paraphrasing AI-generated text or altering details to avoid plagiarism detection systems.

One participant explained that manually editing ChatGPT-generated content, adding personal examples and using paraphrasing tools made assignments appear original, reducing the likelihood of being flagged by plagiarism software.

Another student argued that using ChatGPT during an open-book examination was comparable to consulting a digital textbook, although they deliberately avoided informing the course instructor.

The researchers found that many students prioritised better grades and faster completion of tasks over concerns about academic integrity or the potential erosion of critical thinking skills.

At the same time, they noted that ChatGPT’s growing popularity reflects broader structural challenges within higher education, including large class sizes, limited access to educational resources and insufficient opportunities for personalised academic support.

Calls for AI literacy

Rather than recommending restrictions on ChatGPT, the researchers argue that universities should help students learn how to use generative AI responsibly.

They recommend introducing AI literacy into university curricula, teaching students how to verify AI-generated information, recognise fabricated references and properly acknowledge AI assistance in academic work.

The study also urges educators to redesign assessments so they reward critical analysis, personal reflection and original thinking, making it more difficult to rely solely on AI-generated responses.

Other recommendations include establishing clear institutional policies on acceptable AI use, training faculty members to identify misuse while incorporating AI effectively into teaching, and encouraging open classroom discussions about the ethical implications of generative AI.

The researchers caution that their findings are based on interviews with only 20 students from a single university and therefore cannot be generalised to all Bangladeshi university students. Nevertheless, they argue that the study offers an early glimpse into how rapidly generative AI is reshaping learning habits and academic practices in the country.

As universities around the world continue to grapple with the opportunities and risks of artificial intelligence, the study suggests that the question may no longer be whether students will use ChatGPT, but whether institutions can help them use it responsibly without compromising academic standards.