Infographic dashboard on IIIT Delhi's pilot policy allowing ChatGPT in exams with prompt disclosure required. Shows policy rules, benefits like reasoning evaluation, school implications, and clarification: "Based on media reports & analysis, not official IIIT graphic." Data from educationforallinindia.com Nov 2025.

Dashboard visualizing IIIT Delhi’s ChatGPT policy (AI allowed in exams, submit prompts) and implications for Indian schools.


AI in Indian School Education: Analyzing IIIT-Delhi’s ChatGPT Policy

Abstract

The Indraprastha Institute of Information Technology Delhi (IIIT-Delhi)‘s recent policy permitting the use of AI tools like ChatGPT in examinations and assignments marks a progressive shift in higher education assessment paradigms. By mandating the submission of prompts alongside outputs, the policy emphasises transparency, ethical use, and higher-order cognitive skills over rote memorisation. This review article examines the policy’s implications, drawing on data-driven insights from UDISE+ and Samagra Shiksha frameworks, while aligning with the National Education Policy (NEP) 2020’s vision for technology-integrated learning. It critically assesses the feasibility of extending such AI-assisted evaluations to secondary and higher secondary schools, highlighting prerequisites such as digital literacy and infrastructure. Potential barriers, including the digital divide in rural and marginalised communities, are analysed, and recommendations for inclusive implementation are provided to achieve universal school education by 2030.

Introduction

In an era where artificial intelligence (AI) permeates every facet of human endeavour, educational institutions must evolve beyond prohibition to proactive integration. On December 1, 2025, IIIT-Delhi announced a pilot program allowing students to employ generative AI tools, such as ChatGPT, during exams and assignments, provided they disclose the prompts used to generate responses. This initiative, as articulated by Director Ranjan Bose, reframes AI not as a threat to academic integrity but as a catalyst for responsible innovation:

“Originality will lie in how they frame and refine prompts to obtain meaningful outputs.”

This review situates IIIT-Delhi’s approach within broader Indian educational reforms, leveraging insights from the Education for All in India platform (educationforallinindia.com), maintained by Prof. Arun C. Mehta. With over 247 million school students and persistent enrollment gaps (secondary GER at 78.7% per UDISE+ 2024-25), extending AI to school exams could democratise access to advanced learning tools—provided equity is ensured.

Review of IIIT-Delhi’s Policy: Strengths and Innovations

IIIT-Delhi’s framework is a paragon of adaptive pedagogy, with key features:

  • Transparency Mechanisms: Mandatory prompt submission enables evaluation of reasoning and iterative refinement.
  • Assessment Redesign: Shift from memory to higher-order thinking, aligned with NEP 2020.
  • Ethical Safeguards: Compulsory AI ethics sessions and signed declarations.

Prospects for Extension to Secondary & Higher Secondary Levels

Extension to Classes 9–12 is feasible in the near future. CBSE’s Open Book Assessments (2025–26) and state-level AI literacy pilots indicate readiness. Prof. Arun C. Mehta’s article “Envisioning Indian School Classrooms in 2035: The Transformative Role of AI and Smart Technologies” and “AI in Indian School Education: Challenges, Curriculum & Teacher Training” advocate early integration of prompt engineering into the syllabus from Class 8 onward.

The Digital Divide: A Critical Barrier

UDISE+ 2024-25 data reveals stark disparities:

  • Only 63.5% of schools have functional internet.
  • Rural government schools lag far behind urban private institutions.
  • An estimated 47.4 million children remain out of school, mostly from marginalised communities.

Without aggressive bridging measures (BharatNet, offline AI tools, community devices), AI-integrated exams risk becoming another vector of inequality.

Prerequisites for AI-Assisted Examinations

  1. Reliable digital infrastructure and offline-capable AI tools
  2. AI & digital literacy for students and teachers (including prompt engineering, bias detection, ethical use)
  3. Curriculum integration of machine learning basics and AI ethics
  4. Robust policy framework for transparency, data privacy, and equity monitoring

Conclusion and Recommendations

IIIT-Delhi’s bold experiment can become a national model, but only with deliberate inclusion. By 2030, AI-assisted assessments could help realise NEP’s vision of universal, equitable, and quality education—if India invests decisively in closing the digital divide today.

Key Recommendations

  • Phased optional rollout in secondary boards from 2026–27
  • Allocate ≥10% of Samagra Shiksha funds for rural AI infrastructure
  • Make AI literacy mandatory in pre- and in-service teacher training
  • Establish nationwide longitudinal research via NIEPA/NCERT

Suggested Reading

© 2025 Education for All in India | educationforallinindia.com