Foundation/Mozilla Education/India

Mozilla Education and Responsible Computing in India

Overview

Mozilla's education work in India began in 2023 with the Responsible Computing Challenge (RCC), with the goal of strengthening how ethics, responsibility and the societal impacts of technology are integrated into computing education. Given the scale and diversity of India's higher-education ecosystem, RCC India was designed to work across different kinds of institutions and disciplines, bringing computer science into conversation with the social sciences, humanities, law and design. From the outset, the programme emphasised inclusion, privacy and safety, sustainability, digital public goods and ethical approaches to technology, while encouraging responsible computing to be embedded across the curriculum rather than treated as a standalone or final-year subject. Since then, Mozilla's education work in India has expanded beyond curriculum change to explore how students can become active participants in shaping technology. Through university grants, student projects, student ambassadors, convenings and community engagement, the work has sought to connect what students learn in classrooms with the social contexts in which technologies are ultimately deployed. The current phase builds on these lessons through a focus on Human Scale AI and students as co-creators. Through Education Assemblies, the Human Scale AI Buildathon and university-based programmes, Mozilla is experimenting with models in which students work across disciplines and alongside communities to identify problems, question when AI is appropriate, and build technologies grounded in real-world needs. Together, these efforts aim to develop approaches to AI education where students are not simply taught how to use or build AI systems, but are equipped to critically examine and actively shape the technologies they create.

Responsible Computing Challenge India: Building the Foundation

The first phase of Mozilla's education work in India focused on a fundamental question: how do we ensure that responsibility is not an afterthought in computing education, but part of how future technologists are trained from the beginning? Through the Responsible Computing Challenge, Mozilla supported universities to experiment with different models for embedding ethics, social responsibility and public-interest thinking into computing education. Rather than prescribing a single curriculum, institutions were encouraged to develop approaches suited to their disciplines, students and local contexts. This produced a deliberately diverse portfolio. At IIT Indore, the programme supported the development of five Responsible Computing-themed courses spanning areas including green software, software verification, digital governance and accountability; two were formally incorporated into the B.Tech Computer Science curriculum, with three others progressing through the institutional approval process. At Digital University Kerala, Responsible NLP was integrated into postgraduate computing education alongside flipped-classroom teaching, faculty development and wider institutional outreach. The work also contributed to the creation of a Student Ambassador Programme intended to sustain student leadership around responsible computing beyond the grant period. Other institutions pushed the model beyond the classroom. At Ashoka University, students investigated technology in its real-world context — interviewing gig workers, conducting field research and developing projects around privacy, surveillance, digital identity, misinformation and AI. Several projects continued beyond the course, with external organisations supporting some of them and one project moving towards state-government deployment. At D.J. Sanghvi College of Engineering, a Responsible AI Product Development programme combined 16 weeks of teaching with 44 weeks of mentoring and product development. Six student teams completed the programme, with teams progressing beyond prototypes to products with active users and formal relationships with hospitals, educational institutions and other real-world partners.

What We Learned

Across these different experiments, a common pattern emerged: responsible computing became substantially more meaningful when students moved from discussing ethical principles to encountering their consequences in the real world. Students responded strongly when they could investigate actual communities, users and institutions; build and test technologies; encounter questions of privacy, bias, safety and inclusion themselves; and receive feedback from people outside the classroom. Ashoka's experience was particularly instructive: field engagement helped students avoid what the faculty described as "optimising from afar" — designing solutions for people without actually engaging with them. This became an important lesson for the next phase of Mozilla's education work in India: Students should not only learn responsible technology. They should have opportunities to practise responsibility by investigating, questioning and building technology in the world around them. That insight provides the bridge from Responsible Computing to the current focus on Human Scale AI and Students as Co-Creators.

The India Grantee Portfolio

The first RCC India cohort brought together 10 higher-education institutions across India, each testing a different approach to embedding responsible computing into education. The portfolio deliberately ranged from curriculum reform and accessibility to AI fairness, environmental sustainability, community engagement and student-led product development. Nine of the ten projects have now completed their grant activities; IIT Guwahati remains ongoing.

CHRIST (Deemed to be University), Bengaluru — Responsible Computing Across Disciplines

CHRIST University approached responsible computing as an interdisciplinary educational challenge, bringing computing into conversation with ethics, social sciences and the wider consequences of technology. Its work focused on creating learning experiences in which students could examine technology not simply as a technical artefact, but through questions of responsibility, societal impact and human values. The project also helped demonstrate how responsible computing could move beyond a standalone ethics course and become part of a broader institutional teaching culture.

Indian Institute of Technology Delhi — Responsible Computing through Interdisciplinary Education

IIT Delhi's RCC work explored how responsible computing could be integrated into engineering education by connecting technical learning with questions of ethics, society and public impact. The initiative created opportunities for students and faculty to engage across disciplinary boundaries and helped establish responsible computing as a legitimate area of inquiry within a highly technical academic environment. Its contribution to the cohort was particularly important in demonstrating how India's leading engineering institutions could broaden the definition of what constitutes computing competence.

Indian Institute of Technology Guwahati — Computational Acoustic Ecology

Ongoing

IIT Guwahati is developing an interdisciplinary model connecting computing, environmental stewardship and the ecological and linguistic diversity of Northeast India. The project uses computational acoustic ecology and soundscape data to help students examine how computing systems perceive — or discard — the natural and human environments around them. Its planned outputs include an open-source soundscape database covering urban and non-urban locations in Northeast India, classroom-ready computational acoustic ecology modules, integration across humanities, design, engineering and computer science courses, and dissemination to educators outside IIT Guwahati. This is the only RCC India grantee project still in progress.

Indian Institute of Technology Indore — Embedding Responsible Computing into the B.Tech Curriculum

IIT Indore focused directly on institutionalising responsible computing within undergraduate Computer Science and Engineering. The project developed five new courses spanning applied blockchain accountability, software verification, assisted-living systems, digital governance and green software; two have already been formally incorporated into the B.Tech curriculum, with three progressing through institutional approval. The project also trained more than 55 faculty members from regional institutions, creating a multiplier effect beyond IIT Indore itself.

National Institute of Technology Warangal — CLILuminating Changemakers and TAGS

NIT Warangal combined curriculum innovation with community engagement through its CLILuminating Changemakers project. Responsible computing was integrated into disciplines including Computer Science, Biotechnology, Mathematics and Computing, and Metallurgical and Materials Engineering, alongside faculty-development programmes. Its most distinctive intervention was TAGS — Tech Ambassadors on Ground, which took responsible computing beyond the campus by working with rural and tribal communities, including tribal women, on digital literacy, technology, entrepreneurship and community problem-solving.

Indian Institute of Information Technology Kottayam — Explainable and Responsible AI

IIIT Kottayam concentrated on making fairness, accountability, transparency and Explainable AI (XAI) practical components of computer science education. Students worked hands-on with techniques including SHAP, LIME and fairness-aware model tuning, while interdisciplinary workshops connected technical AI development with ethics and governance. The programme produced student research and practical tools including an XAI Dashboard Prototype for bias detection and model interpretability, while extending responsible-AI conversations through workshops, ideathons and collaborations with institutions including IIT Madras and IIT Hyderabad.

Marian College Kuttikkanam — Equitable Digital Access for People with Disabilities

Marian College built a comprehensive programme around digital accessibility and disability inclusion, including a certificate programme, learning platform, accessibility resources and experiential engagement with people with disabilities. One of its strongest outcomes was WebEase 1.0, a student-developed open-source accessibility widget that received an Indian patent and has been integrated into 33+ websites and used by more than 5,000 people. The project also triggered institutional change: accessibility was incorporated into course-planning templates, the college website was made accessible, and physical infrastructure and library resources were expanded for people with disabilities.

Ashoka University — The New Geography of the Information Age

Ashoka University's project moved responsible computing deliberately out of the classroom and into the field. Students investigated issues including gig work, digital identity, privacy, surveillance, misinformation, AI in healthcare and algorithmic systems through interviews, field visits, data collection and experimentation. Rather than treating the final assignment as the end of a course, students developed "springboard projects" intended to continue afterwards: multiple teams did so, at least two attracted external support, and one moved toward state-government deployment. The project demonstrated the value of preventing students from "optimising from afar" by requiring direct engagement with the people and communities affected by technology.

Dwarkadas J. Sanghvi College of Engineering — Responsible AI Product Development

D.J. Sanghvi tested a particularly builder-oriented model: a 16-week Responsible AI Product Development course followed by 44 weeks of mentoring and guided product development. Twenty-nine students participated in the full programme, six teams completed it, and another 118 students participated in a responsible-computing ideathon and pitch event. Crucially, teams were pushed beyond classroom prototypes toward real users: student ventures secured active users and formal relationships with hospitals, educational institutions and businesses. Projects included Fairify, an all-female team's platform for testing bias in NLP and large language models.

Digital University Kerala — Responsible NLP and Responsible AI Education

Digital University Kerala developed and implemented a structured Responsible NLP module while building a broader responsible-AI learning ecosystem around it. The programme combined curriculum development with flipped-classroom approaches, faculty development, ideathons, expert sessions and outreach across more than 15 academic institutions. Approximately 500 students outside the formal course were reached through outreach activities, while a Mozilla Student Ambassador Programme created a pathway for students to lead responsible-computing activities themselves. Student projects addressed areas including privacy-preserving document processing and hate-speech detection for underrepresented code-mixed Indian languages.

From Classroom Learning to Student-Led Projects

A significant outcome of RCC India has been what happened after responsible computing entered the classroom. Across the cohort, students did not only study questions of fairness, accessibility, accountability and social impact; they began translating those principles into tools, field projects and community interventions. Several projects have subsequently moved beyond coursework into real-world deployment.

Civinc — Student-Built Civic Technology Deployed in Nanjangud

Emerging from Ashoka University's RCC-supported ecosystem, Civinc is a student-built civic technology platform designed to make local government more accessible and accountable. The platform gives citizens direct access to councillors, ward officials and municipal employees responsible for services such as sanitation, roads, property tax and building licences, including through ward-specific QR codes placed around the city. In October 2025, Civinc was formally deployed with the Nanjangud City Municipal Council in Karnataka, making Nanjangud the first urban local body in India to digitise its civic data through this model. The project was conceptualised by Ashoka alumna Aishwarya Sunaad, with its technology developed by students Harsh Raj and Aditya Sinha, and supported by Ashoka University's Isaac Centre for Public Policy and the Mozilla Responsible Computing Challenge.

Web Accessibility Widget — Building Accessibility into the Web

At Marian College Kuttikkanam, student Jerit Baiju developed a free web-accessibility widget aimed at removing the cost barrier associated with commercial accessibility products. The tool supports users with visual, cognitive and physical accessibility needs through features including text and spacing adjustments, high-contrast modes, dyslexia-friendly fonts, link highlighting, screen-reader support and content reading. Designed as a scalable tool, improvements can be propagated automatically to websites already using the widget. Rather than treating accessibility only as something students learn about, the project demonstrates students building accessibility directly into digital infrastructure.

FACTS-H — Making AI Decisions Explainable

Students at IIIT Kottayam developed a suite of Explainable AI dashboards that turn abstract principles such as fairness, transparency and accountability into working technical systems. The FACTS-H Dashboard analyses AI decisions, uses SHAP and feature-importance techniques to explain predictions, identifies potential biases and allows users to examine model-performance metrics. A related AI FORA Dashboard lets users work with real datasets, change input parameters and observe how those changes affect AI predictions. These tools demonstrate how responsible AI concepts can become part of the actual engineering and evaluation workflow rather than remaining a separate ethics discussion.

EyeXAI — Explainable AI for Medical Imaging

The IIIT Kottayam work also produced EyeXAI, an explainability tool for retinal-image classification. Rather than simply returning an AI-generated classification, the system uses Grad-CAM visualisations to show which areas of a retinal image influenced the model's assessment. The project provides a concrete example of responsible-AI principles being applied in a high-stakes domain: students are being trained to ask not only whether an AI model produces an answer, but whether that answer can be inspected and understood.

TAGS — Students Taking Responsible Computing into Communities

At NIT Warangal, responsible computing moved beyond the university through TAGS — Tech Ambassadors on Ground. Students and faculty worked with rural and tribal communities through participatory programmes combining digital literacy, responsible AI and community problem-solving. The programme developed multilingual materials in Telugu and extended classroom learning into workshops with tribal women and youth. One documented iteration engaged 56 tribal women in Chintapalli, many encountering AI tools for the first time, while another worked with 42 rural and semi-urban engineering students to identify community problems and develop technology-enabled responses.

Ai4Sure — Responsible AI for 160 SHG Women Leaders

One of the strongest extensions of TAGS was Ai4Sure, delivered by NIT Warangal in partnership with the Government of Andhra Pradesh in Krishna District. The programme trained 160 Self-Help Group women leaders in AI literacy, digital safety, misinformation, privacy and practical applications of AI for entrepreneurship. Participants experimented with AI for marketing, business records, customer outreach, translation and community problem-solving. Importantly, the programme adopted a cascade model: each participating SHG leader was expected to support another 10–15 women, creating the potential to extend learning to approximately 1,600–2,400 women through community-led dissemination.

Fairify — Testing AI Systems for Bias Before Deployment

Developed by students at Dwarkadas J. Sanghvi College of Engineering, Fairify is a bias-testing platform designed to help developers identify unfair behaviour in NLP and AI systems, particularly in sensitive applications such as hiring, lending and education. Instead of offering a generic fairness score, the platform allows developers to test models against biases specific to their own use cases and demographic contexts. It incorporates approaches including Counterfactual Sentence Testing (CST), Sentence Encoder Association Tests (SEAT), and a distributive-justice framework, while also enabling developers to upload or generate customised datasets for testing. What makes Fairify particularly relevant to RCC is the shift from discussing fairness as an abstract ethical principle to building fairness evaluation into the AI development process itself. The project is also explicitly designed for developers without specialised expertise in algorithmic fairness, potentially lowering the barrier to conducting responsible-AI audits.

Crediboost — Exposing Manipulation in Online Reviews

Another student team at Dwarkadas J. Sanghvi College of Engineering developed Crediboost, a browser-based tool investigating how fake reviews influence consumer behaviour and digital marketplaces. Rather than stopping at fake-review detection, the project seeks to analyse the relationship between suspicious reviews and changes in product popularity, traffic and engagement. Its proposed browser extension surfaces this information directly while a user is shopping, including the proportion of potentially fake reviews, changes over time and their estimated influence on product performance. The project addresses a core responsible-computing question: how can technology help users understand when digital information environments are being manipulated? By making these signals visible at the point where consumers make decisions, Crediboost explores how technical interventions can strengthen transparency and informed choice in online marketplaces.

Engraph — Lowering Barriers to Open-Source Participation

Engraph, also developed through the D.J. Sanghvi student programme, tackles a different form of technological exclusion: poorly documented software. The student team is developing an AI-powered system that analyses software repositories and automatically generates structured documentation covering project architecture, dependencies, configuration, features and visual elements. Its architecture combines repository parsing, dependency graphs, Neo4j and GraphRAG to build a contextual representation of a codebase rather than simply generating a generic README. The responsible-computing dimension is particularly relevant to open source: inadequate documentation creates barriers for newcomers and limits who can meaningfully participate in software development. Engraph therefore uses AI not simply to automate developer work, but to make technical projects more understandable, accessible and collaborative. These projects emerged from a larger Mozilla-supported Responsible Computing Fellowship model run with D.J. Sanghvi College of Engineering and SimPPL. The programme gives students academic credit while taking them through problem identification, literature review, user interviews, customer discovery, co-development partnerships, prototyping, testing and sustainability planning. Importantly, projects are expected to produce practical tools for academia, civil society or local government rather than ending as classroom assignments.

Additional Ashoka University Projects

Trust & Privacy in Digi Yatra

Students investigated privacy, consent and public trust in India's facial-recognition-based airport system through 60+ user surveys, field research at Delhi airport, expert and stakeholder interviews, and an RTI request to the Airports Authority of India. The project surfaced a significant gap between users' privacy expectations and their continued adoption of biometric infrastructure, demonstrating how students can critically examine technologies already deployed at population scale.

AI, Algorithms & Electoral Polarisation

Students analysed political social-media discourse during the 2024 Maharashtra Legislative Assembly election, using RoBERTa-based sentiment analysis on posts and replies from major political parties. The research found that opposition-blaming, communal and emotionally charged content generated more polarised responses, while policy-oriented content produced relatively more neutral engagement. The project combined computational analysis with stakeholder interviews and explored mechanisms for greater platform and electoral accountability.

Childhood Under Surveillance

Students examined the privacy and social implications of Delhi's deployment of approximately 146,800 CCTV cameras across around 700 government schools, including classroom surveillance and proposed parental access to footage. Using India's constitutional privacy framework, the project assessed questions of legality, necessity, proportionality, safeguards, children's autonomy, and the long-term effects of normalising surveillance in educational environments.

Mozilla Responsible Computing Ambassador Program

Building on the university-level work of RCC India, the Mozilla Responsible Computing Ambassador Program was piloted at Digital University Kerala (DUK) to explore a peer-led model for taking responsible computing beyond university classrooms. Seven first-year students were selected through a competitive application and interview process and formally began the programme in August 2025. The ambassadors underwent structured learning covering Responsible AI, Natural Language Processing, privacy, fairness, AI governance and the societal impacts of technology. Their learning included research on global AI regulation and the EU AI Act, case studies on data privacy and gender bias, presentations, discussions and final projects. A central feature of the programme was moving students from learners to educators. The ambassadors helped develop a Responsible Computing curriculum for school students, with separate foundational and advanced modules and adaptations for CBSE and State syllabus contexts. After receiving training in pedagogy and classroom engagement, they delivered workshops under the theme "Safe and Responsible AI for a Responsible Future", combining AI literacy with discussions of privacy, bias, ethics and responsible decision-making. Three workshops reached 220 school students across PM Shri Kendriya Vidyalaya CRPF Pallipuram, SSM Higher Secondary School Chirayinkeezhu, and Kendriya Vidyalaya Pattom. The pilot demonstrated a potential multiplier model for responsible computing education: universities train student ambassadors, ambassadors translate that knowledge for younger learners, and school students begin identifying problems they themselves could address through technology. Early follow-on ideas emerging from the school engagements include a traffic-management system being developed by students of Pattom Government Higher Secondary School and an AI-assisted wristwatch concept for senior citizens proposed by students from KV Pallipuram. The pilot has therefore provided an initial model for expanding responsible computing and AI literacy through student-led outreach, with the longer-term possibility of developing a broader network of student ambassadors across India.

RCC Global Conclave — Kerala, 2024

In October 2024, India hosted the first global conclave of Mozilla's Responsible Computing Challenge, bringing together RCC partners, educators, students, funders and government representatives from India, Kenya and the United States. Held in Kerala from 21–24 October, the conference was organised in collaboration with USAID and the Indian Institute of Information Technology Kottayam. More than 100 participants came together to exchange lessons from RCC programmes and explore how responsible computing could be embedded more deeply into technology education and practice across different geographic and institutional contexts. Under the theme "Responsible Computing for Sustainable Futures," the conclave combined RCC grantee showcases, keynotes, workshops, international discussions and student-led sessions. Participating institutions presented their work; students led a panel and participated in a dedicated feedback session; and educators and programme partners jointly explored the future direction of RCC. Discussions compared responsible computing approaches in India, Kenya and the United States and brought perspectives from academia, government, civil society and international development. The programme also included participation from K. Francis George, Member of Parliament for Kottayam, and a fireside conversation between RCC and USAID leadership. A defining feature of the conclave was the role given to students as participants in shaping the programme rather than simply as its audience. Student perspectives informed discussions about how responsible computing should evolve, while the final day, hosted with IIIT Kottayam, provided participants with a closer view of responsible computing in the Indian context.

The conclave helped connect the work underway across individual RCC institutions into a broader international community and reinforced an emerging principle for Mozilla's education work in India: students should have a meaningful role in shaping how technology is taught, built and governed.

Current Work: Human Scale AI and Students as Co-Creators

Building on the lessons from the Responsible Computing Challenge, Mozilla's current education work in India is moving from teaching responsible computing to creating opportunities for students to actively shape the future of AI. The emerging model centres on Human Scale AI: technologies designed around human and community needs, rather than beginning with the assumption that AI itself is the solution. Students are positioned not simply as users of AI tools, but as co-creators who can question, design, test and build technology in collaboration with educators, communities and other disciplines.

Human Scale AI Buildathon — CHRIST University

The Human Scale AI Buildathon, organised in partnership with CHRIST (Deemed to be University), Bengaluru, is a national interdisciplinary student programme that challenges young people to build AI around problems that matter to people and communities. Rather than beginning with a predetermined technology or asking students simply to demonstrate what AI can do, the Buildathon encourages teams to begin with a real problem, understand its social context, critically examine whether AI is appropriate, and then develop solutions that are responsible, inclusive and grounded in human needs. The programme is designed as a multi-stage journey rather than a conventional short-format hackathon. Students from across India participate through registrations and regional rounds, followed by mentoring and development leading to a national finale at CHRIST University's BGR Campus in January 2027. Shortlisted teams will bring working prototypes to the finale, with winning teams receiving recognition, further mentorship and opportunities to continue developing their work, including a pathway for selected participants to present their projects at Mozilla Festival 2027 in Barcelona. The Buildathon was formally launched at CHRIST University's BGR Campus in 2026, beginning a six-month national programme of student engagement and building. The launch brought together students, faculty and Mozilla representatives through a keynote, panel discussion and the launch of the programme website. More broadly, the Buildathon represents an evolution of Mozilla's education work in India: from asking how responsible technology should be taught to creating spaces where students can demonstrate what responsible, Human Scale AI can look like when they are given the agency to build it themselves.

Community Co-Creation and Human Scale AI — NIT Warangal

Building on the earlier TAGS (Tech Ambassadors on Ground) programme, Mozilla and NIT Warangal are developing a community-grounded Human Scale AI programme that brings university students together with underserved communities to identify challenges and build solutions collaboratively. Instead of arriving in communities with predefined AI use cases, the programme begins with discovery and listening. Students will work alongside tribal youth, Self-Help Group leaders and other community participants to identify problems and opportunities that may otherwise remain invisible to conventional technology-development processes. The programme will create five Community Challenge Briefs, which will become the starting point for five interdisciplinary student teams. Each team will work with a community co-designer and undertake responsible-AI training, design thinking, community immersion and participatory co-design before developing prototypes. The aim is to produce at least three community-validated Human Scale AI prototypes built from the ground up with underserved communities, rather than merely testing university-designed solutions on them. Beyond the individual prototypes, the programme is intended to produce a replicable Community Incubator Framework, a Human Scale AI Learning Report and a Mozilla case study documenting what worked, what did not and what could be scaled. In this sense, NIT Warangal functions as an experiment in a broader question for Mozilla Education: what would AI innovation look like if communities traditionally treated as technology's end users were instead involved in defining the problem and shaping the technology from the beginning?

Human Scale AI Accelerator and Offline AI Tutor — IIIT Kottayam

Mozilla's work with Indian Institute of Information Technology Kottayam takes the Human Scale AI approach into education and low-resource learning environments. The year-long student accelerator brings together students, faculty and schools to explore how AI can support learning while responding to constraints that mainstream AI products often assume away — including reliable internet connectivity, access to computing infrastructure, language diversity, privacy and the realities of rural classrooms. At the centre of the programme is the development of an offline, multilingual and privacy-preserving AI tutor for students in Classes 8–10. Rather than simply adapting an existing chatbot for schools, participating students will combine responsible-AI learning, community engagement, co-design and hands-on prototyping to develop and test the system with educators and learners. The programme is designed to ensure that the technology is shaped by classroom needs and local contexts, while giving university students experience of building AI with the people who will ultimately use it. The programme will also develop faculty capacity, community AI-literacy resources, open learning materials and a model that can be replicated in other low-resource educational environments. Together with the CHRIST Buildathon and NIT Warangal programme, the IIIT Kottayam accelerator represents the next phase of Mozilla Education in India: CHRIST creates a national platform for student-led Human Scale AI innovation; NIT Warangal tests deep community co-creation; and IIIT Kottayam explores how those principles can translate into sustained, deployable technology for education.