Lucknow: The use of AI tools in public administration in India will be effective only when their databases are ‘Indianised’, said experts while addressing a seminar on “AI and New Dimensions of Good Governance,” organised by the public administration department of Lucknow University on Monday.They said that databases should reflect India’s linguistic, cultural, social and administrative realities and align with the Indian value system.Keynote speaker, Prof Chharu Malhotra from the Indian Institute of Public Administration, New Delhi, explained that AI trained on western datasets often failed to understand local governance structures, rural livelihoods, and regional socio-economic dynamics.“Indianised datasets can improve predictive governance, policy analysis, welfare delivery, reduce algorithmic bias and strengthen data sovereignty. Western databases cannot work with the Indian system, they are very different,” she said.Chief guest and vice-chancellor of Maharaja Ganga Singh University, Bikaner, Prof Manoj Dixit said, “Developing indigenous AI models that reflect our administrative realities and social diversity is critical.”He also stressed AI’s role in disaster management, policy evaluation, corruption control and promoting gender inclusive governance.The speakers also pointed out that Indianised datasets can help reduce algorithmic bias, ensure fairness in sectors such as healthcare and agriculture advisories.Principal secretary, IT and electronics, Anurag Yadav and LU vice-chancellor Prof Jai Prakash Saini were present.
