I asked ChatGPT a question: “How many college students ask you to do their assignments or write their research papers on an average day?”
Although it responded by saying that there is no data it can offer, it went on to describe a typical day of students’ interaction with AI as involving everything from writing and rewriting assignments to solving homework, researching topics, preparing for exams, and increasingly, asking the chatbot to complete substantial portions of academic work. One line in its response read: “There absolutely are students who essentially say: ‘Here’s the assignment. You do it. I have to submit it in two hours.’”
There are a few data points from independent research too, that point to this trend, revealing a practice that is rapidly becoming commonplace across college campuses, where the method for evaluating students includes a mix of assignments, projects, and dissertation. Although school students, especially tech-savvy middle and high school students seem to also be using AI, a large percentage of schools seem to rely on traditional handwritten examinations to test students’ learning; making it difficult for students to use AI.
At the heart of the issue lies the burden on teachers, who are now expected not only to teach, assess, and mentor, but also to determine whether a student’s work is genuinely their own, constantly adapt assignments to make them less susceptible to AI-generated responses, and find new ways of engaging students in lessons.
The challenge, teachers say, presented itself in 2019. Recalling the first such incident, C Anupama, assistant professor, Department of English, Women’s Christian College, shares, “There is this story by Jayakanthan titled ‘Trial by Fire’ and there was a girl who registered her research topic on it. But her entire term paper was on another author’s book, maybe it had a similar title, but she submitted it and we found out that she had used AI.” The student then hadn’t realised that AI can misunderstand information or rely on incorrect data and generate responses.
Archna Prasad, principal of MOP Vaishnav College for Women, asserts that students’ dependency on AI has grown exponentially since after the Covid-19 pandemic. “It is quite funny that students don’t realise that teachers can identify the language of AI. Especially because, today teachers are aware of AI and are using it as a tool to help make notes for class or integrate updated concepts. So, when students’ assignments or a thought paper is written in the same language that AI uses, it becomes very evident.”
While teachers admit that they cannot expect students to not use AI completely, they flag concerns with students not engaging critically with concepts and developing a genuine understanding of the ideas even if they may be generated by AI. Deepthi V, assistant professor, Department of Journalism, Madras Christian College, says, “Not just one or two, but almost everyone I was mentoring for research had used AI to generate their research ideas last year. When I asked them to explain their topic, they weren’t able to. They were struggling. Because they didn’t understand the topic itself, what they submitted was also copy-pasted from AI.”
These challenges have compelled teachers to rethink and adapt the ways in which they assess students today. Apart from colleges and universities including AI detectors in their courseware, these teachers have had to individually create alternative ways of assigning homework.
To encourage students to write by themselves, Anupama says that students have now been mandated to write their term papers during class hours, allowing them the time to research, with or without the aid of AI, at home. Deepthi adds that many of the assignments have been made more practical, where students are asked to go out on the field, or stage a play, etc., instead of submitting lengthy, written assignments. Colleges too are mandating courses on prompt engineering, on responsible AI usage, and more, so as to empower teachers with a better understanding of technology, hoping that they would impart the same values in students.
Despite their best efforts, teachers observe that students seem to have an “AI addiction” as they are relying on AI for everyday communication. While it may seem harmless, teachers note that this growing dependence can sometimes prevent students from developing the confidence and skills needed to communicate independently.
Deepthi recalls one such incident where a student had asked a chatbot to draft a letter to the ‘Dean of Student Affairs’, except the chatbot addressed the letter to the ‘Dean of Affairs’. The student hadn’t even noticed the error. “When we saw it and asked him, he admitted that he had used AI and that he had used the microphone to prompt. It didn’t pick up the correct designation. What is shocking is that this student didn’t even re-read what AI had given him. He blindly submited it,” she explains.
Anupama too, laments that the biggest misconception about AI among students is believing that AI cannot make a mistake. She comments, “The mistakes made by AI are atrocious but students strongly believe that AI is right. Students’ very thinking process is challenged today. They are not even trying, to be honest.” Deepthi corroborates, adding that students’ thinking ability today is at an all time low. “This makes students completely unaware of their own abilities or limitations. As a result, they are losing confidence in themselves by the time they graduate.”
Archna believes that, rather than saying students’ ability to think has diminished, it is more accurate to say that they have become increasingly unwilling to think. “AI throws a bunch of data at them, but the ability to rationalise and determine whether that information is right or wrong, that is where AI has spoiled students. They don’t want to think, almost like they are lazy,” she says. And that, the teachers unanimously agree, is the challenge they are now navigating, a problem more psychological than academic.
