Five years ago, India launched its most ambitious education reform in decades. The National Education Policy (NEP) 2020 set out a powerful roadmap to move away from rote memorisation toward real understanding, from content delivery to competency, and from standardised instruction to personalised learning. But the success or failure of this initiative will not be defined by how many children we reach, but by how well they learn. Outcomes, not just access, are the true test of this policy.
Since its introduction, we have made encouraging progress. As per the ASER 2024 report, enrollment levels have shown significant improvement across age groups and regions in rural India. Student attendance increased from 72.4% in 2018 to 75.9% in 2024. The National Curriculum Framework for School Education (NCF-SE) has been launched. There is greater awareness that education must be student-centred, application-based, and inclusive. But even as these structural shifts begin to take root, the fundamental issue remains unresolved: Is this increased access to education translating into effective learning?
The data, while showing marginal improvements across the years, suggests we still have a long way to go. The ASER report shows that while 84% of rural households now have smartphones, only 42% of Class 5 students can read a Class 2 text, and approximately 30% can solve basic division problems. Results of the National Achievement Survey (2023) that tests proficiency in Language, Maths, Environmental Studies, Science, and Social Studies, indicate that only 42% of students grade 3, 33 % of students grade 5, 27% of students grade 8, and 22% of students grade 10 are at proficient and advanced levels of competency.
This disparity in improvement of enrollment vs learning outcomes suggests that India’s education challenge is no longer just about reach. It’s about efficacy. Despite greater access to digital infrastructure, online learning content, and evolving curricula and teaching practices, we still don’t have enough clarity on why student learning hasn’t seen a much greater change.
The private sector, too, is following a similar pattern. Edtechs have brought lessons to remote geographies, delivered tutoring at scale, and made exam prep more affordable. But success has traditionally been measured through metrics like daily active users, session time, and course completions. This mismatch between what we measure and what we value has left a generation overexposed to content but under-equipped in cognition.
Beyond metrics to outcomes
Slow growth isn’t a policy problem. This isn’t a technology problem either. It is an implementation problem. We have confused digital access with learning progress, and now we must reassess how we define the accomplishments of our tech-led offerings.
Measuring educational outcomes requires different tools from measuring user engagement. It demands sophisticated assessment frameworks that can capture conceptual understanding, not just factual recall. It requires longitudinal tracking that follows students’ intellectual development over months and years. And diagnostic tools that pinpoint exactly where a student is stuck and not just what they got wrong.
NEP 2020 already provides a viable blueprint with its emphasis on competency-based learning, formative assessments, and flexible instructional design. What we need now are systems and tools that can implement this vision effectively at scale, and without overburdening teachers.
This is where AI, used wisely, can be transformative
Instead of simply digitising textbooks or recording lectures, we can now build platforms that dynamically adapt to a student’s pace, identify areas of confusion, and provide targeted interventions. These systems can track learning not by time spent, but by mastery achieved.
For example, compact, subject-specific AI models, when trained on real curriculum data and student performance trends, can outperform larger, generic models, offering precise feedback, remediation, and revision support. In high-stakes competitive exams, where the margin of error is razor-thin, such models have shown early promise in helping close conceptual gaps faster and more effectively.
Tools can be used to allow teachers to conduct instant in-class polls, track attendance, and run real-time quizzes. These real-time diagnostics give educators immediate insight into where students are struggling without needing additional grading or paperwork.
But measuring learning requires more than just technology. It requires a mindset shift. We must move from tracking activity to monitoring progress. This is why we propose building what might be called ‘learning outcomes as a service’, a model where content, instruction, assessment, and intervention are built around clear learning goals. Such systems would leverage existing and new assessment infrastructure to check subject proficiency, use adaptive technology to customise learning, and rely on continuous analysis to ensure target goals are met.
Implementing this will require partnerships across the education ecosystem. Schools, NGOs, policy experts, and technology firms must come together to design solutions that prioritise conceptual clarity, lateral thinking, and creative problem-solving. Formative data that already exists—on classroom performance, exam trends, student behaviour—can be used more intelligently to inform remediation. But that will only happen if we set a new benchmark for both public and private educational institutions and organisations.
But is there a collective will to put outcomes ahead of optics, and learning ahead of legacy? India stands at a crossroads. We can continue to find hope in marginal improvements and innovation that is limited to pilot projects. Or we can take the harder path of systemic change, creating educational ecosystems that prioritise student learning. The stakes couldn’t be higher. In a global economy driven by innovation and problem solving, we are not just talking about education, we are talking about India’s future competitiveness in a knowledge-first economy.
Published – July 29, 2025 02:16 pm IST
