Abstract
The rapid integration of Artificial Intelligence (AI) and Large Language Models (LLMs) into India’s educational landscape is significantly ahead of the development of corresponding policy and governance frameworks. This paper provides findings from an original primary survey of 125 students across three age groups (below 12, 13–19, and 20 and above) in Bangalore, India, examining the frequency, nature, and student perceptions of LLM use for Science, Technology, Engineering, and Mathematics (STEM) learning. The results show us an alarming rate of AI usage that demands immediate attention: 48% of respondents use AI for academic purposes everyday, 38.4% use it multiple times per week, and only 3.2% reported no AI use.
Survey findings also reveal that students rate AI’s teaching effectiveness highest in computer science (3.82/5) and mathematics (3.65/5); that freedom from fear of judgment, rather than just convenience, is the strongest motivator for AI use; that 92.8% of respondents expect AI to become a permanent tutoring presence or support in schools and colleges; and that there is significant student resistance to AI-based assessment. These findings collectively reveal a governance gap in which student-led AI usage for learning purposes has grown suddenly without any regulatory frameworks, accuracy standards, age-specific safeguards, or academic integrity guidelines.
This paper uses these empirical findings to identify where regulatory action is most urgently needed in India’s existing policy landscape including the National Education Policy 2020 and the Digital Personal Data Protection Act 2023, and to make relevant, targeted policy recommendations addressed to the Ministry of Education, UGC, and relevant regulatory bodies.
Keywords: Large Language Models, Artificial Intelligence, STEM Education, AI Governance, Educational Policy, Student Survey, EdTech, NEP 2020, Digital Personal Data Protection Act, AI in Education
Read the complete paper. Download the PDF from the top-right corner.
