IIT Kanpur's AI Tool Can Identify Welfare Beneficiaries Without Surveys

The Indian Institute of Technology (IIT) Kanpur has announced that an AI-powered system it developed can help governments identify households requiring welfare support, using data that is already available within government systems.

The system, called Family Score, was created in collaboration with the Andhra Pradesh government. The first phase of the project tested whether AI could serve as a reliable substitute for traditional household surveys when assessing socio-economic conditions.

According to the institute, the findings show that an ensemble of AI models can outperform conventional surveys in determining a household's socio-economic status, while requiring minimal additional cost and effort from the government. The approach eliminates the need for fresh data collection, making it faster and more cost-effective for identifying beneficiaries at scale.

The findings were presented to Andhra Pradesh Chief Minister N Chandrababu Naidu and senior officials by an IIT Kanpur team comprising Director Manindra Agrawal, Dean of the Wadhwani School of AI & Intelligent Systems Nitin Saxena, and Manish Srivastava, CEO of Airawat Research Foundation.

Discussions during the meeting focused on developing a roadmap to improve the living standards of low-income families through technology-enabled, data-driven interventions. The members also deliberated on building a comprehensive citizen database to better understand family-level needs, and on leveraging AI to identify the means, needs, and eligibility of families for more targeted welfare delivery.

Other topics included strengthening data-driven decision-making to make welfare schemes more transparent and outcome-oriented, and exploring technology-driven solutions to help the government connect families with appropriate schemes. Naidu reportedly appreciated the system's capabilities and its potential to reach the right families with targeted support.