An exploratory study undertaken by IIIT-H’s Human Sciences Research Center (HSRC) has won the Best Paper Award at the 6th India Public Policy Network Conference for analysing how public procurement processes are quietly influencing the development and governance of AI systems in India. 

When India talks about governing Artificial Intelligence (AI), discussions often revolve around future regulations, policy frameworks and ethical guidelines. But where are some of the country’s most important AI governance decisions being made?

A research team from IIIT-H’s Human Sciences Research Centre (HSRC) comprising Prof.  Aakansha Natani, Assistant Professor at HSRC, Siddhi Wadekar, PhD scholar, and Sujal Deoda, a dual-degree student pursuing BTech in Computer Science and MS in Computing and Human Sciences (CHD), explored precisely this question in their award-winning paper, ‘Emerging Institutional Pathways for AI Governance in India: Evidence from Public Procurement and Outsourcing.’ The paper received the Best Paper Award (Practice Track) at the India Public Policy Network (IPPN) Conference 2026 hosted by the National Law School of India University (NLSIU) on 8-11 June 2026.

The conference is one of India’s leading forums in the field of public policy. This year’s edition was organised in collaboration with University of Cambridge, UK, National Council of Applied Economic Research (NCAER), International Public Policy Association (IPPA), and several other partner institutions.

Emerging From Ongoing Research
The study grew out of an ongoing research project at HSRC examining critical policy gaps in India’s AI and data infrastructure. “Unlike the EU, India currently does not have a dedicated AI Act. Instead, AI governance is guided through advisory documents, frameworks and guidelines; and this regulatory landscape is still evolving in India and many other developing countries,” says Siddhi. That reality led the team to an important research question: if formal regulation is still evolving, where are governance practices actually emerging from?

Studying the Language of Governance
To answer this question, the researchers turned to publicly available government tenders related to AI systems. The idea arose from a simple observation. Across India, central and state governments are increasingly adopting AI-powered systems, from citizen services and administrative tools to law enforcement applications. Yet most government agencies do not possess the technical infrastructure or specialised expertise required to build or manage these systems internally. As a result, they often turn to private companies and startups.

“The government needs AI-driven technologies but it might not have the necessary capacities and expertise to build and maintain it. So it often resorts to outsourcing these services through tenders,” explains Sujal. Every such outsourcing process generates a trail of documents and requirements – tenders, procurement notices, technical specifications, eligibility requirements, contractual obligations and compliance conditions. The team’s methodology involved analysing patterns, terms, requirements and standards embedded within tender documents issued by state and central government agencies.

Methodology
Their approach combined computational analysis with qualitative policy analysis, allowing them to study both the language used in procurement documents and the broader governance implications of that language. The goal was not simply to understand what governments were procuring, but to examine how procurement documents articulate expectations around AI systems and what those expectations might reveal about emerging governance approaches. As the researchers describe it, the project sought to understand whether procurement itself is becoming a mechanism through which AI governance is being delegated, negotiated, contested and implemented in India.

Early Insights from an Emerging Field
The researchers note that the study revealed the importance of procurement and outsourcing as central features of India’s AI ecosystem. The work highlights how governance discussions cannot be limited to formal regulations alone. In rapidly evolving technological domains, standards, requirements and accountability mechanisms may also emerge through operational processes such as procurement.

As Prof. Natani points out, researchers need to understand not only the technological specifications of AI systems, but also how policy documents, procurement processes, regulations (or lack of regulation), and institutional practices translate into governance framework. “For this kind of research, you need computer science expertise to understand the technical specifications required for AI systems, as well as political science expertise to examine what’s being written and what is not being written in policy and procurement documents” she says.

The Importance of Interdisciplinary Research
Perhaps the most remarkable aspect of this work is the collaboration that made it possible. The project brought together expertise spanning computer science, computational social sciences, public policy and political science. Such intersections are increasingly necessary for studying complex socio-technical systems like AI, where technical decisions often carry significant societal implications. According to Prof. Natani, this kind of research environment is a distinctive strength of IIITH. “These sort of studies are possible at IIITH and Human Sciences Research Centre specifically because of our dual degree programme” she says.

The institute’s academic ecosystem allows students and researchers to move across disciplinary boundaries. In this project, a Computer Science student pursuing an integrated B.Tech and MS by research in Computing and Human Sciences worked alongside a PhD scholar in Political Science to address questions that cannot be answered through a single disciplinary lens.The collaboration pointed towards a broader pattern indicating that governance may not be solely shaped by laws passed in legislatures but also through everyday mechanisms by which technologies are commissioned, procured and deployed.