Study Buddy, TA, Or Outsourced Brain? What’s The Future Of AI In Indian Education?

When it comes to AI, people fall on the spectrum of techno-utopian optimism (AI will solve all problems) versus calls for outright boycotts of the hallucination engine. 

“I think that we've been forced into this dichotomy of being in one camp or the other,” Tarang Tripathi, co-founder of Aawaaz Education Services, an education support organisation building an AI tool for teachers, told The Secretariat. “We lie somewhere in the middle of optimism and pessimism, but curiosity centers us,” says Tripathi. 

Whether one is starry eyed or looks on with squinting suspicion as prompts run, AI is here to stay. At least for now. 

India, with the largest student population on ChatGPT in the world, is fast becoming a test case for how AI can shape education policy and classroom practice. 

From Accelerator To Action Plans

Just last month, OpenAI announced its Learning Accelerator in India. It’s the first such programme anywhere in the world. The tech giant will provide five lakh free ChatGPT licences for six months to students and teachers across schools, universities, and research institutions. 

Alongside this, AICTE (All India Council for Technical Education) has partnered with OpenAI to distribute an additional 1.5 lakh licences specifically to technical institutes.

Earlier this year, AICTE declared 2025 as the Year of AI, signalling a push to embed AI across the higher education ecosystem. Plans include capacity building programmes for faculty, and collaborations with companies like Perplexity, which has agreed to provide AI access to 40 million students across 14,000 institutions.

Curricula for engineering and management have already been updated with AI-focused courses.

The National Education Policy (NEP) 2020 turned five this year. It laid the foundation for using AI and other emerging technologies to transform learning and build an inclusive, tech-enabled education ecosystem.

But policy momentum is not enough on its own. “AI is only as good as the data it is trained on, and that data is not neutral,” Tripathi cautioned. “We need frameworks that put guardrails in place so that biases don’t end up hardcoded into classrooms,” he said. 

The question is also what happens when these licenses expire. An OpenAI representative told The Secretariat that while the company did not have any further updates to share at this stage, they would look into it in the near future. 

A Tool, Not A Replacement  

The key, however, is viewing AI as a "tool," not a "foundation". “The goal is not just to do things faster, but be able to do something that you previously were not able to," says Shaivi Soni, teacher and education researcher. “AI should be treated as a tool to help students and teachers do more, not as a replacement for human judgment. The machine suggests, the teacher decides,” she said. 

Aawaaz's own tool, "TeachAIde," is an example of this philosophy. Built by researchers and teachers, TeachAIde is a hyper-contextualised application designed to produce differentiated lesson plans that cater to individual student preferences and academic needs. 

Built as a wrapper on OpenAI’s LLM, the tool incorporates anonymised student information gathered through pen-and-paper surveys. It also considers teacher preferences. This level of personalisation allows teachers to build better student relationships, helping them navigate learning more effectively.

Bridging Access And Ethics

While there is a promise of AI in classrooms, experts caution that access remains uneven. The digital divide persists. Students in rural areas or low-income households often lack devices or reliable internet connectivity that could reinforce educational inequities. 

“The challenge is not only to introduce AI tools, but also to ensure that all students can benefit from them,” Tripathi said.

Privacy is another concern. AI systems often require data about student learning patterns, and while Aawaaz anonymises this information, the broader ecosystem must address how sensitive data is collected, stored, and used. Without clear guidelines, there is a risk that student information could be misused or misinterpreted.

Soni emphasises the importance of context in using AI. “AI can augment learning, but it needs to be guided by ethical practices and strong oversight. Teachers still need to interpret, correct, and contextualise what the AI suggests,” she said.

The Road Ahead

While AI in education can be oversimplified to cheating on assignments, Soni says the trick lies in teaching students to acknowledge where they are using the tool. 

Challenges like AI hallucinations, where models generate inaccurate information, is crucial. While OpenAI claims ongoing fixes experts say that hallucinations may never be entirely eradicated. 

That’s why digital literacy for both teachers and students is integral. Students, particularly through K-5, don’t possess the cognitive understanding to critically evaluate what an LLM spits out. That’s why there’s a need for age-appropriate exposure and training. 

Licences, curriculum updates, and policies provide a framework, but their impact will be limited without teacher training, ethical safeguards, and measures to reduce inequities. As Tripathi puts it, “AI is a powerful amplifier of what already exists, so we need to make sure it amplifies curiosity, inclusion, and fairness, not bias or inequality.”

AI in Indian education shouldn’t be about replacing teachers or textbooks, but personalising learning, lightening cognitive load, and creating spaces where students engage more deeply with knowledge. 

Tripathi says he envisions classrooms as "creative laboratories" where students experiment freely, and teachers, empowered by AI tools that reduce administrative burdens, can dedicate more time to nurturing individual student relationships and fostering a holistic learning experience.

“AI is pretty cool,” the Class X student said. Despite concerns, many educators agree.