The New Academic Assistant
Generative Artificial Intelligence is no longer a futuristic concept; it's a daily reality on campuses across India. Recent reports show a significant number of students use AI tools for everything from homework help to exam preparation. One report found
that education and learning account for nearly 20% of AI conversations in India, with homework making up a substantial portion of that. For many students, using AI is becoming a first instinct, not a last resort. These tools can draft essays, solve complex math problems, and write code, offering an unprecedented level of assistance that blurs the line between a helpful tool and an academic crutch.
Beyond a Simple Shortcut
It’s easy to frame this as a simple issue of cheating, but the reasons students turn to AI are often more complex. Many are grappling with immense academic pressure, tight deadlines, and a feeling of being overwhelmed. For some, AI can feel like a necessary tool to manage workloads and reduce stress. The issue is also pedagogical. When assignments are perceived as unengaging or focused on rote memorisation, students may not see the value in completing them without assistance. If an AI can successfully complete a task, it begs the question of whether the assignment was truly measuring critical thinking or merely the ability to synthesise information—a skill technology has now mastered.
The Educator's Impossible Task
While students navigate this new landscape, educators are caught in a difficult position. The primary response has been a technological arms race, with a boom in AI detection software. However, study after study has shown these tools to be unreliable. They are prone to both false positives—incorrectly flagging human work as AI-generated—and false negatives. This unreliability poses a serious ethical problem, creating a culture of surveillance and mistrust that can damage the student-teacher relationship. Relying on flawed detectors risks unfairly penalising students, particularly those who are English language learners or have writing styles that deviate from the norm. Even the makers of ChatGPT shut down their own detection tool due to a lack of accuracy.
Rethinking the Rules of Engagement
Given the limits of detection, many educators and institutions are realising that the solution isn't to ban AI, but to adapt to its presence. Some universities in India are starting to formulate official policies, though many are still applying general academic integrity rules to this new context. The most promising path forward appears to be rethinking assignment design. This includes a shift towards assessments that AI cannot easily complete, such as in-class presentations, oral exams, project-based work, and assignments that require personal reflection or engagement with specific in-class discussions. Some educators are also finding ways to incorporate AI ethically, teaching students to use it as a brainstorming partner or editing tool while demanding full disclosure of its use.
