By -Vansh Aggarwal

Abstract:

This article explores how AI is reshaping public policy in India, focusing on welfare schemes like MGNREGA and PM-JAY. Drawing lessons from DigiYatra, FASTag, and CoWIN, it examines both the efficiency gains and the risks of privacy loss and bias. The article also proposes integrating AI with Aadhaar to build smarter, citizen-centric welfare delivery.

Introduction:

Artificial Intelligence (AI) has become one of the most significant forces shaping governance and public policy today. India, with its massive population and ambitious welfare programs, has begun integrating AI into public schemes in different ways. The push for a digital economy, combined with the government’s flagship initiatives like Digital India, has laid down the groundwork for adopting AI in governance. NITI Aayog, in 2018, also came up with a National Strategy for Artificial Intelligence, calling it “#AIforAll”, with the focus areas being healthcare, agriculture, education, smart cities, and smart mobility. The government has set up platforms like IndiaAI to encourage both public and private stakeholders to develop AI solutions. The main motivation behind this shift is to deal with the scale and complexity of India’s welfare schemes. For example, India has over 1.3 billion Aadhaar IDs, over 700 million internet users, and multiple central and state schemes that require data management at an unprecedented scale. AI becomes attractive because it can process patterns in this data faster than human systems ever could.

AI current landscape in India:

Several schemes already showcase how AI and related technologies are being applied. DigiYatra, which was launched in December 2022, uses facial recognition technology to allow seamless air travel without having to repeatedly show boarding passes or ID cards. Another case is FASTag, which uses RFID and AI-powered systems to enable automatic toll collection. The CoWIN platform during the COVID-19 pandemic is perhaps the most well-known example. It was not just a booking portal but used AI-enabled scheduling to manage vaccine slots, track supplies, and even prevent fraudulent registrations. Similarly, the Pradhan Mantri Fasal Bima Yojana (PMFBY) has started using AI-driven satellite imagery and predictive weather modelling to assess crop damage and decide insurance payouts more accurately. 

From these examples, some clear lessons have emerged. AI-backed systems improve efficiency, reduce delays, and increase transparency. However, there are also concerns. Facial recognition in DigiYatra has raised privacy questions. FASTag has left out some citizens who are not digitally literate or banked. CoWIN, while a success, exposed India’s digital divide, where many rural citizens struggled to access vaccine slots. Similarly, in PMFBY, small farmers without smartphones or internet access risk being excluded from benefits. This means that while AI adoption in governance has promise, it must be coupled with strong safeguards around privacy, inclusivity, and accessibility.

AI Integration in MGNREGA:

The Mahatma Gandhi National Rural Employment Guarantee Act (MGNREGA) is one of the world’s largest public works programs, providing guaranteed employment for rural households. While the scheme has provided livelihood security for millions, it has struggled with several problems. Ghost workers and fake job cards remain an issue, leading to leakages. There are often long delays in wage payments due to poor fund allocation. Further, demand for work is highly seasonal, but forecasting is weak, resulting in a mismatch between the availability of work and demand. AI can play a transformative role here. Predictive models could be used to estimate rural job demand by analysing migration data, crop cycles, and even weather forecasts. 

For example, areas expecting poor rainfall could be prioritised for generating more employment under MGNREGA, since demand will naturally rise. Fraud detection is another potential use. Similarly, fund allocation can be made more efficient by AI systems that predict demand and automatically recommend optimal fund transfers to states and districts, ensuring timely wage payments. However, safeguards are essential. First, AI systems should not be allowed to automatically exclude workers without human review. Second, grievance redressal mechanisms must be transparent so that workers wrongly flagged can appeal. Third, strict data protection norms must be followed to ensure that rural workers’ personal information is not misused.

AI Integration in Ayushman Bharat – PM-JAY:

Ayushman Bharat, i.e., Pradhan Mantri Jan Arogya Yojana (PM-JAY), is India’s flagship health insurance scheme, covering over 50 crore beneficiaries. It promises free treatment up to Rs 5 lakh per family per year in empanelled hospitals. However, the scheme faces several challenges. Insurance fraud is common, with cases of hospitals making false claims or inflating bills. Detecting regional healthcare demand in real-time is difficult, which affects planning.  AI can provide strong solutions here as well. Fraud detection can be strengthened using machine learning models that analyse claim patterns to identify suspicious behaviour. For example, if one hospital is consistently claiming a large number of expensive procedures compared to others, it can be flagged for audit. 

Predictive analytics can also help forecast regional healthcare demand. For instance, analysing weather and sanitation data can help predict dengue outbreaks, allowing the government to allocate resources in advance. Telemedicine platforms supported by AI can triage patients, guiding them to the appropriate level of care and reducing the burden on hospitals. Personalised healthcare alerts could be sent to beneficiaries based on their medical history and risk factors, helping with preventive care. But again, policy safeguards are crucial. AI systems must not discriminate against rural or poor patients, who already face barriers in healthcare. Health data is extremely sensitive, so strong privacy protection is needed, especially as India has now passed the Digital Personal Data Protection Act, 2023. 

Broader Policy Ecosystem:

The experiences from DigiYatra, FASTag, and CoWIN provide valuable lessons for welfare schemes like MGNREGA and PM-JAY. DigiYatra shows how seamless biometric authentication can simplify processes, which could be adapted to welfare beneficiary verification. FASTag demonstrates the possibility of automated, real-time financial transactions, which could be replicated in MGNREGA for instant wage payments. CoWIN’s ability to scale during the vaccination drive provides a blueprint for PM-JAY in terms of health data management and predictive allocation of medical resources. The key insight is that India already has functioning proof-of-concept models of AI in governance. The challenge is extending them to welfare schemes while ensuring inclusivity and trust.

Aadhaar already acts as the central identity layer for most welfare schemes in India. Integrating AI with Aadhaar can open new possibilities. AI-based anomaly detection could be used to find duplicate or fake Aadhaar-linked entries across multiple schemes. Real-time predictive checks could prevent overlapping benefits or misuse. For example, if the same individual is drawing subsidies from different overlapping schemes, the system could flag it. AI-powered grievance redressal chatbots could be linked with Aadhaar to allow beneficiaries to raise issues in local languages and receive quick responses. Fraud detection in direct benefit transfers (DBT) could also be enhanced by AI models that flag abnormal withdrawal patterns. Independent AI audits should be conducted to check for bias and misuse. Finally, citizens should have opt-out or consent-based models to maintain trust in the system.

Conclusion:

India today stands at a crossroads where the integration of AI in public policy could redefine welfare delivery. The experiences of MGNREGA and PM-JAY highlight both the immense opportunities and the serious challenges that come with such a transformation. When implemented with care, AI has the power to cut leakages, improve efficiency, and bring essential services closer to citizens. But if safeguards are neglected, the same technology could deepen inequalities, compromise privacy, and leave the most vulnerable behind. Initiatives like DigiYatra, FASTag, and CoWIN prove that India has the technological capability to scale AI solutions effectively. The bigger challenge lies in ensuring that Aadhaar-linked and AI-enabled systems remain citizen-centric, transparent, and fair. The real question, then, is not whether India can lead the world in AI-driven governance, but whether it can do so without sacrificing the rights and trust of its people.

Author’s Bio:

Vansh Vijay Aggarwal is a B.A. LL.B. student at Jindal Global Law School and a columnist at CNES.

Image Source : India AI Mission to focus on deep startups, Inclusive AI