Nagaland University - Lumami Campus.
Lumami, November 2 (MExN): A research team led by Nagaland University (NU) has brought together scholars from Indian institutions to explore how artificial intelligence can support the preservation of India's rich cultural heritage, with the findings published in a peer-reviewed journal.
The study, published in Acta Innovations, focuses on manuscripts, endangered languages, oral traditions and historical records, many of which are at risk due to aging, damage or the fading of traditional knowledge. The team includes researchers from institutions in Tamil Nadu, Delhi, Uttar Pradesh, Telangana and Karnataka, a press release received here stated.
The researchers developed a framework to evaluate how AI tools can be used in preservation, with the goal not just to digitise items but to ensure that deeper meanings, historical accuracy and cultural context are maintained. They stressed that AI should not replace understanding or community involvement, and should be deployed alongside experts and local communities to preserve heritage while respecting ownership and tradition.
The paper was co-authored by Prof Ravi Ranjan Kumar, NU; Major Dr N Kavitha, Vellalar College for Women (Autonomous), Tamil Nadu; Nisha Sengar, PGDAV College Evening, University of Delhi; Dr Sonia Gouri, ABES Engineering College, Ghaziabad; Dr J Lalitha Sridevi, Hyderabad; Kumar Sai Deepak B, Sapthagiri NPS University, Bangalore; and Dr P Deepavathi, Joy University, Tamil Nadu.
Highlighting the need for institutions across India to collaborate on research, Prof Jagadish K Patnaik, Vice-Chancellor, NU, said the initiative reflects the growing potential of AI in safeguarding India's rich and diverse cultural heritage. He added that equally important is ensuring such technological interventions respect cultural contexts, community rights and ethical considerations.
Elaborating on the research, Prof Ravi Ranjan Kumar, Hindi Department, NU, said a key contribution of the study is a framework that evaluates AI-assisted cultural preservation across nine areas - digitisation accuracy, text recognition, classification, translation, visual reconstruction, descriptive information, cultural context, accessibility and ethics. The framework combines technical assessments with reviews by subject experts and relevant communities.
He noted that technical accuracy does not necessarily guarantee cultural accuracy, as an AI system may correctly transcribe a document or translate spoken words while missing historical references, cultural nuances or the significance of a tradition. Expert verification and community consultation are therefore essential, particularly when working with less commonly used languages, unusual scripts, damaged documents and oral histories.
The framework suggested that any AI use must include input from the people whose heritage is being preserved, and calls for guidelines on data use, privacy and access.
The research identified five key areas where AI could support cultural preservation: printed and handwritten documents; photographs and artworks; oral histories and endangered languages; archaeological and architectural heritage; and folklore and community knowledge. Potential applications include making historical documents searchable, organising image collections, identifying damage to artworks, transcribing oral recordings, translating multilingual materials and creating digital maps and three-dimensional models of heritage sites.
The study also highlighted the importance of distinguishing between restoration and reconstruction, recommending the retention of original digital files, documentation of AI-generated changes and clear indication of uncertainty so that users can distinguish original evidence from computer-generated interpretations.
The proposed framework could be particularly relevant to India, where historical manuscripts, regional languages, oral traditions and community knowledge represent a vast and diverse cultural record. In the Northeast, it could help explore approaches to preserving endangered languages and documenting oral histories, provided the tools are suitable for the languages concerned and communities participate in decisions about how their heritage is recorded and shared.
For libraries, museums, universities and archives, searchable digital collections could improve access for students and researchers while reducing the need to repeatedly handle fragile original documents, the researchers said.
The authors recommended piloting the framework across different collections, scripts and dialects to assess its practical value, documenting recognition errors, incorporating expert and community feedback and establishing whether AI-generated outputs are appropriate for their intended uses.
