When Linus Torvalds released the Linux kernel in 1991 with the now-famous Usenet post “just a hobby, won’t be big and professional like gnu,” he could not have anticipated that Indian developers would one day rank among the top contributors to the project that runs on over 90% of the world’s servers. India’s journey in open source – from the early Kernel hackers of the late 1990s to the BharatGPT consortium of 2024 – is one of the least-told technology stories in a country that celebrates its software exports.
Indians in the Linux Kernel: A Presence That Predates the Startup Boom
The Linux kernel maintainer hierarchy is one of the most meritocratic technical communities in the world – there are no corporate mandates, no nationality quotas, and contributions are accepted or rejected purely on technical quality and adherence to the kernel coding standards. Indian developers have carved out a real presence in this system.
By the Linux Foundation’s 2020 Linux Kernel Report, the top contributing organizations to the kernel included Intel, Red Hat, Google, IBM, and Samsung – but embedded in those corporate numbers are thousands of individual contributors, many of Indian origin. Atul Sharma, Anilkumar Konatala, Prashant Bhole, and dozens of others appear across kernel git logs maintaining subsystems including ARM architecture support, networking drivers, and filesystem code. India’s contribution to the 5.x and 6.x kernel series has been particularly visible in the networking and storage subsystem patches.
The parallel story in Estonia is instructive. Estonia’s approach to digital infrastructure has always been open source first – the X-Road data exchange layer that connects all government services is open source (MIT license) and has been adopted by Finland, Namibia, and Azerbaijan. When Estonia needed a national digital backbone, they built it on Linux and open standards. India’s IndiaStack – Aadhaar, UPI, DigiLocker, ONDC – follows a similar philosophy, but with critical differences in licensing and governance that are worth examining.
IndiaStack: Open Source Policy Infrastructure at Scale
India’s digital governance transformation – from the world’s largest biometric identity system to IndiaStack – is arguably open source contribution to the world – not as individual software packages, but as a policy architecture that has influenced digital identity programs in over 50 countries. The core APIs are publicly documented, the UPI protocol specification is open, and the ONDC (Open Network for Digital Commerce) protocol is designed as an interoperability standard rather than a proprietary platform.
UPI processed 13.89 billion transactions in January 2025 alone, according to the National Payments Corporation of India (NPCI). The protocol that enables this scale is open – any bank or fintech company can connect to the NPCI infrastructure and offer UPI services. This is fundamentally different from how Visa and Mastercard operate (proprietary networks with licensing fees) and closer to how Estonia’s X-Road or Sweden’s BankID work – interoperable, standards-based, and accessible.
The National Informatics Centre (NIC) has maintained open source infrastructure for Indian government services since the 1980s. GIGW (Guidelines for Indian Government Websites) mandates open standards for government websites. But India’s open source policy record is inconsistent – the government still procures large amounts of proprietary software (Microsoft Office is standard in most government offices) while simultaneously building open digital public infrastructure.
“The IndiaStack architecture is perhaps the most consequential contribution India has made to the global digital commons – a reference model for how a developing country can build sovereign digital infrastructure without depending on proprietary platforms.”
OECD Digital Economy Outlook 2023, Chapter on Digital Public Infrastructure
BharatGPT, Sarvam AI, and Krutrim: India’s Open AI Moment
The most significant development in Indian open source in the 2020s is the emergence of indigenous AI models, some of which are being released with open weights. This matters because the global AI infrastructure is currently dominated by proprietary models (OpenAI’s GPT-4, Anthropic’s Claude, Google’s Gemini) with closed weights and no access to training data.
BharatGPT, a consortium led by IIT Bombay and Corover.ai with support from the Ministry of IT, was announced in 2023 with the goal of building a large language model trained on Indian languages and data. The project explicitly commits to open weights for academic research use. By 2024, the consortium had trained models on 22 scheduled Indian languages, though full open release timelines remain subject to compute and governance decisions.
Sarvam AI, founded in 2023 by former Google Brain researchers Vivek Raghavan and Pratyush Kumar, has taken a different approach. Sarvam released Sarvam-1, a 2-billion parameter model trained predominantly on Indian languages, as fully open source under the Apache 2.0 license in 2024. This is significant: Apache 2.0 allows commercial use, modification, and redistribution, making Sarvam-1 genuinely accessible to Indian startups, researchers, and government agencies without licensing fees.
Krutrim, founded by Ola’s Bhavish Aggarwal, announced a 7-billion parameter model trained on Indian language data in 2024. Unlike Sarvam, Krutrim has been more cautious about open weight release, positioning itself as a commercial offering. The contrast between Sarvam and Krutrim illustrates a fork in Indian AI strategy: open-weight public goods versus proprietary commercial models.
| Model | Organization | Languages | License | Status (2025) |
|---|---|---|---|---|
| Sarvam-1 | Sarvam AI | 11 Indian languages | Apache 2.0 (fully open) | Released |
| BharatGPT | IIT Bombay + Corover.ai | 22 scheduled languages | Research open | In development |
| Krutrim | Krutrim AI | Multiple Indian languages | Proprietary | Commercial release |
| AI4Bharat IndicBERT | IIT Madras | 12 Indian languages | MIT License | Released |
AI4Bharat, the IIT Madras project that predates the current wave, deserves particular mention. Since 2020, AI4Bharat has released IndicBERT, IndicBART, IndicNLPSuite, and several speech models (IndicWav2Vec, Seamless M4T fine-tunes for Indian languages) – all under MIT or similar open licenses. This is the kind of foundational open source infrastructure that enables hundreds of downstream applications.
Indian FOSS Communities: FSCI, Software Freedom Law Centre, and the Grassroots
India has a strong if under-resourced free and open source software community. The Free Software Community of India (FSCI), founded in 2012, is a volunteer network that advocates for software freedom in government procurement, digital rights, and education policy. FSCI has run campaigns against mandatory Aadhaar-based biometric authentication for school mid-day meal schemes and has successfully pushed for open standards in government e-procurement systems.
The Software Freedom Law Centre India (SFLC.in), founded by advocates Mishi Choudhary and Prasanth Sugathan in 2011, provides legal support to free software projects and digital rights cases. SFLC.in has filed interventions in right-to-be-forgotten cases, net neutrality proceedings before TRAI, and privacy cases that affect open source developers in India.
Kerala remains the most significant state-level open source success story. Kerala’s IT policy has included open source mandates since 2001, and the IT@School project – which deployed Linux-based computers in government schools across the state – ran for over a decade before being partially replaced by Android tablets. The project demonstrated that government procurement can choose open source at scale when there is political will.
The Gap: Where India Falls Short of South Korea and Germany
South Korea’s open source track record offers a useful comparison. The Korean government’s Open Source Software (OSS) portal, maintained by the National Information Society Agency (NIA), tracks open source adoption across government agencies and provides a procurement framework that defaults to open source when technically equivalent to proprietary alternatives. By 2022, the Korean government had migrated over 40% of its internal desktop systems to open source operating systems (NIA, Korea Open Source Annual Report 2022).
Germany’s Munich and Schleswig-Holstein governments have both undertaken LiMux-style migrations (and in Munich’s case, reversed course back to Windows) – but the process of attempting open source migration at city scale generated institutional knowledge that has informed European Union digital sovereignty policy, culminating in the EU’s Open Source Software Strategy 2020-2023 which mandates open source publication of publicly funded software.
India has no equivalent national open source strategy document. The government’s approach to digital infrastructure has been pragmatic rather than principled – open where it makes economic sense (UPI, DigiLocker) and proprietary where incumbent relationships persist (most desktop software, most ERP systems). A coherent national policy framework, similar to what France enacted with the Loi pour une République Numérique (2016), would close this gap.
What Every Indian Can Do: Five Levels of Citizen Action
- Personal level: Switch at least one tool in your daily workflow to an open source alternative. LibreOffice for Office documents, Firefox for browsing, Signal for messaging, GIMP for image editing. File your income tax returns on the government portal directly (which runs on open infrastructure) rather than through private intermediaries. Each choice creates a market signal that open source tools can meet professional standards.
- RWA/building level: If your housing society uses proprietary software for billing, maintenance requests, or communication, propose an open source alternative. Many RWA management tasks can be handled by self-hosted tools like ERPNext (Indian-developed, globally used) or Nextcloud for document sharing. Pilot one tool per quarter.
- Ward/local body level: File RTI requests asking which software your municipal corporation uses for tax collection, property records, and complaint management. Ask whether these systems are built on open standards that allow data portability. Push for open API access to municipal data (property records, zoning maps, complaints data) – several cities including Bangalore and Pune have begun releasing this data but access remains inconsistent.
- City/state level: Advocate for an open source first policy in state government IT procurement. Reference Kerala’s IT@School experience as a successful model. Write to your state IT secretary asking for a published procurement policy that considers open source alternatives before proprietary ones. Several states have such policies on paper – ask for the enforcement record.
- National level: Support a “Public Money, Public Code” mandate for software developed with central government funding. The EU’s JOINUP platform for sharing government open source is a model India could adopt. File RTI requests asking how many MeitY-funded software projects have been released as open source. Contribute to AI4Bharat or FSCI if you have technical skills. Attend public consultations on the Digital India 3.0 framework and submit comments supporting open source mandates.
The Lever That Closes the Gap
India’s open source story in 2026 is a story of enormous latent capacity and insufficient policy structure. The individual talent exists – Indian developers contribute to every major open source project. The institutional infrastructure exists – IndiaStack is a genuine world-class contribution. What is missing is the connective tissue: a national open source strategy that creates incentives for government agencies to release publicly-funded software, a procurement framework that defaults to open source, and an educational curriculum that teaches computer science students to contribute to open source communities rather than just consume commercial cloud services.
South Korea achieved its open source adoption rate through sustained policy pressure over 15 years. Estonia built its digital sovereignty on open standards through a national strategy that started in 1997. India’s Rs 6,000 crore Quantum Mission demonstrates that the government can commit to frontier technology at scale – open source infrastructure deserves a similar commitment. India has the talent. It needs the policy framework to turn individual contributions into national infrastructure.
Explore India’s Tech and Digital Infrastructure Stories
India’s digital transformation covers far more than open source software. Read about how India built the world’s largest biometric identity system and the quantum mission shaping the next frontier of Indian computing.