#155 A sneak peak into the Indian Government's thinking on AI
In this edition of Technopolitik, Shobhankita Reddy summarises key takeaways from a recent parliamentary committee report on Artificial Intelligence
A new parliamentary committee report, presented in March 2026, provides a good overview into the Indian government’s thinking about Artificial Intelligence (AI).
It refers to AI as a “kinetic enabler to leapfrog traditional development barriers and catalyse large-scale socio-economic transformation in the country”. In the tussle between competition and diffusion in a sector increasingly framed by geopolitical contestations, the Indian government’s priorities are clear.
Agriculture: The Kisan e-Mitra chatbot provides real-time scheme information in 11 regional languages, handling over 8,000 queries daily. Additionally, AI-based monsoon advisories have been piloted to assist 3.8 crore farmers across 13 states.
Education: Initiatives like Shiksha Setu in Assam have used AI to record attendance, successfully identifying and removing 4 lakh ghost students. Furthermore, the FutureSkills PRIME program has already trained over 1.68 lakh individuals in AI-related courses.
Employment: AI is projected to create 47 lakh new tech jobs by 2027.
The report provides several examples of how different government ministries and departments are using AI.
As for AI in the military, the report provides several use cases where it is under experimentation or deployed, such as - surveillance and intelligence gathering, language translation on the edge, equipment design, development and maintenance, specialized applications such as cyber threat detection and robotic systems.
But the following challenges are mentioned -
Strategic GPU Dependence: India currently has a strategic dependence on imported accelerators (GPUs) that are critical for AI, which the government aims to break through indigenous initiatives.
Data Constraints: There is a gap in the availability of standardized and “learnable” datasets necessary to build reliable military AI models.
Legacy Systems: Older platforms (tanks, aircraft, and artillery) were not built for AI integration, making upgrades complex and expensive.
Verification Risks: AI/ML techniques are not always amenable to verified decision-making, which may result in unintended outcomes in critical operations.
And specifically, on lethal autonomous weapon systems, the report says this -
One major challenge as submitted by the Ministry is the lack of consensus on how to define the autonomy of weapon systems. A standard definition that accounts for levels of autonomy could help guide an incremental approach to proposing limits. AI ML based techniques are not amenable for verified decision making and hence are likely to result in unintended outcomes as well. The Committee would like to call upon the Ministry to apprise them of the way-out to the challenges that are posed in this field.
Challenges and constraints identified for India’s adoption and implementation of AI are -
Infrastructure Accessibility: There exists a notable disparity in the accessibility of advanced AI infrastructure, particularly between urban and rural regions. Accessibility not only pertains to physical resources but also extends to user interfaces, which must be user-friendly and designed for individuals with varying levels of technical expertise.
Cost Barriers: The high costs associated with advanced computing resources hinder broader participation in AI initiatives, particularly among startups and small businesses that may lack the financial resources for substantial investment in AI technology.
Shortage of Skilled Workforce: A significant challenge is the scarcity of individuals who possess the requisite skills and expertise in AI technologies and data science. This shortage undermines the effective optimization and deployment of AI solutions.
Data Privacy and Security: The assurance of robust data privacy and security measures is imperative for establishing trust among users and for compliance with existing regulations. Navigating the complexities of data governance remains a significant challenge.
Scalability and Future Readiness: As AI technologies evolve and the volume of data increases, there is a pressing need to develop scalable AI infrastructures that are adaptable to future demands. This includes addressing the requirements for both computational power and data management practices.
Institutional Frameworks: The establishment of a comprehensive institutional framework for data governance, as proposed through initiatives such as the National Data Governance Policy (NDGP) and the National Data Management Office (NDMO), is still in its formative stages. These frameworks are essential for ensuring orderly data collection, management, and access, which are critical for maximizing the potential of AI.
Safe and Ethical use of AI: There is a need for responsible AI frameworks for wider adoption of AI in the Indian context and ensuring solutions developed are relevant linguistically and culturally to India and its citizens.
Some recommendations of the report are -
Fostering talent. Skilling initiatives to mitigate job displacement and social inequality feature highly. Setting up 570 Data and AI Labs in Tier 2 and Tier 3 cities to provide hands-on exposure to tools and datasets.
Preventing AI misuse in financial frauds, intimidation and deepfakes, with a specific mention to the recent attempts at platform governance that aimed to reduce time to action for flagged content from 36 hours to 3 hours, mandate explicit labelling and metadata embedding of synthetically generated content.
Exploring age restrictions for certain social media platforms to safeguard citizens from emerging AI-driven risks.
Expect to come across more news chatter around deepfake regulation, platform governance and “AI for India-Specific Regulatory Framework”, as well as more social media restrictions for the under-age, in the coming months.
Do check out the report in full.
And while you are here, do check out the Takshashila Institution’s brand new offering. An Expert Capsule Course on the Politics and Policy of Critical Minerals.
If you are interested in understanding the complexities at the intersection of geopolitics, technology and economics of critical minerals, then this course is for you. Early-bird application deadline ends April 12th.
Link here.
