On February 18 2026, Central Square Foundation (CSF) launched the second edition of the Bharat Survey for EdTech (BaSE) 2025, a first-of-its-kind national survey that brings out the voice of end-users from low-income contexts and their needs, challenges and perceptions regarding the adoption of education technology and AI. The survey was conducted across 10 states, reaching out to over 12,500 households with children enrolled either in government schools or affordable private schools as well as 2,500 teachers in the same communities.

We caught up with Prof. Ambrish Dongre, Associate Professor, Ravi J Matthai Centre for Education Innovation, Indian Institute of Management (IIM) Ahmedabad and Academic Advisor to BaSE 2025 and 2023. He speaks to us about the need for BaSE 2025, as well as key insights from the survey’s design and deployment. 

1- As the Academic Advisor for the survey, why is the Bharat Survey for EdTech 2025 important? 

Over the last few years, on the heels of the pandemic, we are seeing strong tailwinds in the space of education technology. India’s most recent National Education Policy, NEP 2020, also places an emphasis on the use of technology as far as education is concerned. Following this, many state governments have implemented  large-scale technology interventions. However, despite this vibrancy, there are very few EdTech solutions catering to and scaling in low-income segments. One of the key drivers of this gap is the absence of information regarding the needs and barriers of the end-users in low-income settings. There is a lack of comprehensive, ground-up data and reliable insights on the behaviours, needs and concerns of these users, where tech-based learning could play an important role. It is critical to understand the enablers and barriers faced by households in the low-income sector in order to build equitable and responsive solutions and policies. As innovation in and access to generative AI solutions become increasingly ubiquitous, this imperative is even more important.

This is the context within which we set out to conduct the Bharat Survey for EdTech. The survey seeks to bridge information asymmetry in the EdTech ecosystem by bringing the voice of the low-income to the fore. The hope is that insights from this data will be used by policymakers, EdTech practitioners, product companies and philanthropies to inform policy and fund products and EdTech implementations across the country. As the second edition of this biennial exercise, the survey intends to build upon the insights from its first edition, bringing in a focus on AI to enable timely, data-based decision-making in this emerging sector. 

2- This is the second round of the survey. As BaSE returns to the field since its inception in 2023, what does this edition enable us to understand that was not possible earlier? 

Three key things set this edition apart from BaSE 2023 – scale, scope and thematic focus. 

On scale, we have been able to expand the respondent base of the survey from its first edition. In BaSE 2023, we covered around 6,000 households across 6 states, focusing on a sample size of around 1,000 households per state. In BaSE 2025, the sample was  expanded to include four new states, covering a total of 10 states, taking into account linguistic and regional considerations. We also increased the number of respondents we spoke to, covering a sample of over 15,000 respondents across 10 states. 

On scope, we wanted to get a more holistic view of how EdTech is being used both at home and in the classroom. This year, we expanded the scope of the survey to include the voice of teachers across the country. BaSE is one of the first surveys in India to cover teachers’ access to technology, awareness of and usage of EdTech and AI, as well as teacher sentiment on using technology for teaching and learning. In addition, focus group discussions were conducted across states with adolescents, parents and teachers to supplement the quantitative findings with more context and insights.

Finally, we also expanded the thematic focus of the survey to include timely perspectives on generative AI. The AI module of the survey covers awareness and usage of AI, drivers of AI usage, use-cases for AI, as well as sentiment and perception of risk with respect to the use of AI for teaching and learning. 

3- India is a geographically and linguistically diverse country. How did you ensure that BaSE 2025 captures these differences and is representative of India’s population? 

For the study, we covered 10 states across India, chosen to reflect both digital access and population diversity, while also ensuring a good geographical spread. We used a simple 3×3 matrix that looked at internet penetration and state population size. States were grouped by high, medium and low internet penetration using 2024 data published by the Telecom Regulatory Authority of India (TRAI) and similarly by their share of India’s population using projections based on Census 2011 from the National Commission on Population, Ministry of Health and Family Welfare, Government of India. From across these categories, we then selected 10 representative states: Gujarat, Kerala, Madhya Pradesh, Maharashtra, Mizoram, Odisha, Tamil Nadu, Telangana, Uttarakhand and Uttar Pradesh. We also covered both rural and urban segments within these states.

This process of state selection allowed us to pick states that represent the geographic and linguistic diversity of India, while also capturing variation in digital access and population. Together, this ensured that the study findings reflect a wide range of regional contexts and are more broadly representative across the country. 

4- 15,000 respondents is a large number. How did you balance scale with quality to ensure the integrity of the survey findings?

For a survey of this scope and scale, it was critical to maintain the rigour and reliability of the data. Building upon our learnings from the first edition, we implemented a robust end-to-end quality control framework before, during and after data collection.

Before we got to the field, extensive preparatory work went into creating survey protocols, sampling and field protocols. The tool then underwent two rounds of field testing, enabling us to make further refinements and improvements. Further, during enumerator training, we placed emphasis on not only understanding the questionnaire and sampling protocols, but also on building a shared understanding of what we mean by EdTech and AI in the context of teaching and learning, using practical examples and scenarios. This helped enumerators accurately interpret responses, probe effectively and engage with respondents more confidently, ultimately strengthening the quality and consistency of the data collected.

During data collection, field implementation was closely monitored via on-the-go training, supervisor oversight, geolocation and survey-duration checks and verifications to confirm interviews were conducted correctly. We also regularly reviewed the sample to ensure it aligned with the intended design, benchmarking key demographics against external sources such as the Census projections and data from the Unified District Information System for Education (UDISE).

After data collection, multiple validation checks, eligibility and logic checks were carried out to ensure consistency and accuracy and early trends were tracked for key indicators such as device access and EdTech usage to sense-check results as they emerged.

This process reflected close collaboration among multiple actors involved in the process, including field enumerators, supervisors and our technical partners, working in close coordination to uphold high standards of data quality, accountability and methodological rigour.

5- Who are the different stakeholders that can benefit from BaSE 2025 findings and how can these insights be used most meaningfully?

We hope that the insights and data from BaSE 2025 will be useful for a range of actors in the EdTech ecosystem. Policymakers can use the insights to shape more grounded digital education policies and target public investment where access or usage gaps are largest. EdTech companies and innovators can better understand how learners and teachers are actually using technology, what features they value and where AI tools could add real impact. Educators can use the findings to design more relevant interventions and blended learning models, while researchers can leverage the data to study trends in access, adoption and outcomes over time. Finally, funders and philanthropies can use the evidence to make more informed decisions about what solutions to support and scale, especially in low-income and low-resource contexts.

The survey is an ambitious endeavour to systematically capture the experiences and perspectives of parents, children and teachers on technology in education and translate them into actionable insights for the ecosystem. By bringing together data on access, usage, sentiment and emerging technologies such as AI, it aims to inform effective policy, better product design and programs that truly respond to end-user needs.