Vintage-style map of India's language landscape with floral accents and cultural illustrations.

Preserving Endangered Languages: India's Remarkable Journey

India stands as one of the world’s most extraordinary examples of linguistic diversity. With over 1,600 recorded mother tongues and dialects, and 22 constitutionally recognized languages, the country’s linguistic landscape is as vibrant as it is complex. The Census of India 2011 identified 121 languages spoken by more than 10,000 people each - 22 scheduled and 99 non-scheduled - while historical records from the 1961 Census documented over 1,500 distinct languages. In the educational domain, official data from UDISE+ 2020-21 shows that 28 languages are currently used as mediums of instruction in primary schools, reflecting a deep-seated commitment to multilingual education.

This constitutional and institutional framework - rooted in the Eighth Schedule, which grants scheduled languages official status and protection - is a testament to India’s commitment to linguistic pluralism. These languages are used across administration, education, media, and culture, and the government has consistently worked to document and preserve non-scheduled languages through various initiatives. Yet, beneath this vibrant surface lies a pressing crisis.

The Challenge: A Race Against Time

Despite its rich heritage, India is home to the highest number of endangered languages in the world, according to UNESCO. The government’s Scheme for Protection and Preservation of Endangered Languages (SPPEL), launched in 2013 under the Central Institute of Indian Languages (CIIL), Mysuru, has identified 117 endangered languages in its first phase - languages spoken by fewer than 10,000 people. Beyond these, 81 tribal languages are classified as vulnerable, and 42 are critically endangered. The urgency is palpable: every time an elder passes away without passing on their mother tongue, an entire linguistic universe risks extinction.

In recent years, however, a wave of innovation - driven by artificial intelligence, digital archiving, and grassroots community participation - has begun writing a new chapter in India’s language preservation journey.

The New Frontier: AI, Digital Archiving, and Community Action

🤖 AI-Powered Translation and Preservation

At the forefront of this movement is Adi Vaani, India’s first AI-powered translator for tribal languages, launched in beta by the Ministry of Tribal Affairs in August 2025. Developed by a consortium led by IIT Delhi in collaboration with BITS Pilani, IIIT Hyderabad, and IIIT Nava Raipur, Adi Vaani is more than a translation tool - it is a national mission to digitize, preserve, and revitalize tribal languages.

Available on the Play Store and via a dedicated web platform, it currently supports Santali, Bhili, Mundari, and Gondi, with Kui and Garo under development. The platform enables real-time text and speech translation between Hindi/English and tribal languages, and features Optical Character Recognition (OCR) for digitizing manuscripts and primers. Crucially, it combines AI with community-driven approaches - Tribal Research Institutes, experts, and communities are actively involved in data collection, validation, and iterative development.

This initiative is being further strengthened by platforms like BHASHINI (India’s National Language Technology Mission) and BharatGen, which provide multilingual support across governance, healthcare, and education. The BhashaDaan initiative, a citizen-driven effort under BHASHINI, encourages community-led contributions of speech, text, and image data to build robust AI language models for classical and endangered languages like Pali, Prakrit, and Avesta Pahlavi. Digitized language data from the SPPEL and Sanchika (CIIL’s digital language repository) further enrich AI model training for these multilingual solutions.

📜 Government-Led Documentation at Scale

SPPEL remains the backbone of institutional preservation. The scheme aims to document the language and culture of endangered mother tongues by developing primers, bilingual/trilingual dictionaries (in both electronic and print formats), grammatical sketches, and pictorial glossaries. Community members serve as language consultants during fieldwork, ensuring authenticity. SPPEL is working toward documenting approximately 500 lesser-known languages in the future.

Additionally, Tribal Research Institutes across states have been preparing bilingual dictionaries, trilingual proficiency modules, and primers for students from Class I to V under Multi-Lingual Education (MLE) interventions aligned with the New Education Policy 2020. In Odisha alone, primers have been developed in 15 tribal languages, and CIIL has developed primers for 117 languages in collaboration with NCERT.

💾 Grassroots Digital Archiving: The Assam Model

A powerful example of community-led preservation unfolded in Assam in January 2026, when three critically endangered languages - Khamyang, Tai Phake, and Singpho - were digitally preserved under an initiative led by the Nanda Talukdar Foundation with support from the All Assam Students’ Union (AASU). The project carried special urgency: Bhogeswar Thomung, aged 84, is believed to be the sole surviving fluent speaker of Khamyang. The digital archive includes manuscripts, rare texts, historical photographs, and extensive audio recordings documenting vocabulary, pronunciation, oral traditions, and everyday language use, following internationally accepted endangered-language preservation standards, including UNESCO guidelines. AASU leaders have stated that similar efforts will be extended to other endangered languages of Assam in the coming years.

🏛️ New Institutions and Cultural Recognition

In a landmark move for cultural preservation, Union Home Minister Amit Shah inaugurated ‘Shabdalok’ (Museum of Word)  - India’s first immersive language museum - at the National Library in Kolkata in July 2026. Described as a significant initiative to preserve India’s linguistic heritage, the museum aims to expand to encompass digital technologies and AI, enabling Indian languages to evolve with the technologies of the future.

Additionally, the Ministry of Culture, through its seven Zonal Cultural Centres, has been documenting vanishing art forms, folklores, and oral traditions in audio, video, and written formats. The National Inventory of Intangible Cultural Heritage (ICH) maintained by Sangeet Natak Akademi now includes unique elements like the whistled language of Meghalaya. The Sahitya Akademi has been bestowing Bhasha Samman to writers and scholars working in unrecognized and tribal languages, and has established dedicated centers for oral and tribal literature in Agartala and Delhi.

📱 Community-Centric Innovations

A 2025 study from KISS Deemed University highlights that mobile applications, digital storytelling, AI-based linguistic tools, and online archives created by communities are becoming agents of change. Elders are turning into digital storytellers, children into interactive learners, and native speakers into creators and controllers of their linguistic futures. In one remarkable effort, researchers have created ELR-1000, a rich multimodal dataset of 1,060 traditional recipes crowdsourced from rural communities across Eastern India, spanning 10 endangered and under-represented languages - demonstrating that preservation can be as flavorful as it is vital.

Navigating the Pitfalls of AI-Driven Preservation

However, this technological revolution is not without its risks. A blind embrace of AI could inadvertently replicate colonial dynamics or flatten cultural richness in new digital forms. Here are the critical pitfalls that must be acknowledged and addressed:

1. 🎭 Cultural Appropriation and Data Colonialism

Large tech corporations may extract linguistic data from indigenous communities to develop commercial products, without the communities benefiting from or having control over how their languages are represented. When language data is treated merely as a resource to be mined rather than a living cultural heritage, it constitutes “data colonialism.”

2. 📉 Exacerbating Linguistic Inequalities

AI models are overwhelmingly trained on dominant languages like English. Endangered languages lack sufficient digital corpora, meaning AI risks further marginalizing them—reinforcing rather than bridging global linguistic power imbalances.

3. 🧩 Loss of Cultural Context and Nuance

Language carries history, identity, and worldview. AI struggles to capture idioms, proverbs, and culturally specific expressions. The result may be technically correct language that is culturally hollow - a pale imitation of the living tongue.

4. 👻 “Hallucinations” and the Spread of Misinformation

AI systems often confidently generate false information. In endangered languages, this could mean inventing non-existent words, fabricating spurious “cultural teachings,” or producing inaccurate translations, which can actively undermine genuine revitalization efforts.

5. 🔄 Model Collapse and Cultural Inbreeding

When AI models are increasingly trained on AI-generated content rather than authentic human data, they lose linguistic diversity over successive generations. Rare grammatical structures and specialized vocabulary gradually vanish - a form of digital inbreeding that can drive a language toward digital extinction.

6. 🤝 Insufficient Community Participation and Data Sovereignty

Effective preservation must be community-led. Without the full participation, informed consent, and ongoing oversight of native speakers, projects are prone to failure. Communities must have sovereignty over their linguistic data - the right to decide how their language is represented, stored, and used.

7. 💰 The Digital Divide

Many indigenous and rural communities lack reliable internet connectivity, smartphones, or digital literacy. This paradox means that those most in need of preservation tools are often the least able to access them.

The Way Forward: AI as a Collaborative Ally, Not a Replacement

Recognizing these pitfalls is not an argument against using AI - it is a call for greater responsibility. The true promise of AI lies in serving as a collaborative cognitive ally, not a replacement for human effort. To achieve this, AI-driven projects must be:

- Community-centered: Designed and led by the communities whose languages are being preserved, serving their priorities rather than external agendas.
- Respectful of data sovereignty: Ensuring communities retain full control, ownership, and decision-making power over their linguistic data.
- Committed to cultural authenticity: Prioritizing long-term, sustainable cultural transmission over quick technological fixes.
- Inclusive of the digital divide: Building offline capabilities and providing training to bridge access gaps.

India’s Adi Vaani initiative offers a promising model in this regard. By actively involving Tribal Research Institutes, linguistic experts, and community members in data collection, validation, and iterative development, it ensures that technology respects and reflects cultural knowledge. Initiatives like BhashaDaan similarly empower citizens to contribute their linguistic heritage on their own terms.

A Vibrant Future

The recognition and protection of India’s diverse languages reflect the country’s commitment to linguistic pluralism - but the work is far from over. With 117 languages already being documented under SPPEL, new digital archives being built in Assam, AI tools bringing tribal languages into the digital age, and immersive museums celebrating the written word, India’s linguistic heritage is not just being preserved - it is being given a vibrant future.

The goal is not simply to digitize languages into sterile archives. It is to create the social, educational, and technological conditions where languages thrive naturally - passed down from grandparents to grandchildren, spoken in markets and homes, sung in folk songs, and whispered in prayers. For those of us who love languages and want to keep them alive in our own lives, even something as personal as a daily language journal is a small act of preservation — a commitment to living in a language rather than just studying it. When technological development is paired with community empowerment and ethical governance, AI becomes a powerful guardian of human linguistic diversity. By embracing this balanced approach, India sets an inspiring example for the world on how a multicultural, multilingual society can celebrate its differences as a source of strength, identity, and shared humanity.

Sources: 

Documenting India’s Endangered Languages


📓 Keep a language alive — starting with your own.

The Multilingual Me journal gives you 60 days of guided prompts to write, reflect, and grow in your target language. Every entry is a small act of preservation.

Shop the Journal →
Back to blog

Leave a comment