Whyfores and wherefores of the Centre for Digitisation, AI, and Society
The Computer Science department at Ashoka University has a solid foundation to tackle novel questions around safe and ethical design and deployment of such technologies, write Prof. Subhashis Banerjee and Prof. Debayan Gupta
Computers and networks now mediate a tremendous amount of our daily activity, ranging from shopping to health care to political engagement, and even to friendship and romance. This trend is expected to continue and accelerate for the foreseeable future. India is at the forefront of this wave, especially when it comes to public service applications. Our vast public digital infrastructure enables a bewildering array of public and private sector applications, from the critical to the comical.
The Computer Science department at Ashoka has the unique advantage of being situated amongst strong departments of economics, sociology, political science and philosophy, and people with deep expertise in public health and epidemiology, as well as digital health. As such, we have a solid foundation to attack novel questions around the safe and ethical design and deployment of such technologies. Digitisation and AI provide powerful ways to improve people’s lives, but come with dangerous pitfalls for those who try to use them as blunt-force tools: if one has a hammer as seductive as AI, everything does indeed start to look like a nail.
Digitisation must come with nuance: many attempts at building large public services datasets like national identity systems, health registries, national population and voter registries, public credit registries, income and tax registries, etc. have been questioned on fairness, privacy and other ethical grounds. The concerns have invariably been related to the need for protective safeguards when large data integration projects are contemplated, and acknowledgement of the bias, exclusion, discrimination, privacy, and security problems that these may create. In some situations, they have even had to be abandoned altogether as they were unable to deal with these risks.
In these contexts, digitisation and the use of AI often come with a generic public-good motivation: that somehow, having digitised data and acting upon it in a “data-driven” fashion will automagically result in many good things and few bad ones. This hopeful theory of public good is insufficient for a country as large and diverse as India, where even corner cases may comprise millions (and these are often the most marginalised). Opaque, large-scale systems which use AI may leave citizens feeling helpless to counter decisions: without sufficient use case and outcome analysis, people can and will fall through the cracks.
This is both critical and urgent: as we scale up the use of Aadhaar and UPI, and deploy new infrastructures like ABHA (a national healthcare ID), ONDC (a unified e-commerce and marketplace), and PM Gati Shakti (a national planning and logistics hub), etc., we must push for principled design and planning, not band-aid solutions after the fact.
The Centre for Digitisation, AI, and Society at Ashoka aims to:
- Develop a greater understanding of the technical, legal, economic, social and ethical complexities created by particular uses of technology in large public-facing applications.
- Produce effective frameworks and guiding principles for the design of efficient, safe and trustworthy systems for large public-service applications through research, partnerships, workshops, and consultations with academicians and practitioners from governments and industry.
- Support well-identified research problems to study issues related to digitalisation and society from multidisciplinary perspectives.
- Hold recurring interactions with policymakers, governments and civil society, participate in public discourse through editorials and media interactions, and advocate for wide recognition of the issues identified by research efforts and for the comprehensive adoption of the standards for trustworthiness developed by affiliated researchers.
- Create a platform for interdisciplinary partnerships from within the Ashoka and outside, by facilitating collaborative projects, workshops, conferences and other activities.
- Train students to critically think about these issues so that they may become prepared to participate in and shape the digitization and AI application endeavours mindful of the ethical implications.
We are launching this year with a call for interdisciplinary research projects which align with our vision: we intend to seed these ideas, enabling researchers to start building projects they would not have otherwise started. Our hope is that with this initial impulse, we will help researchers apply for larger-scale funding, facilitate greater conversation and cross-pollination of ideas across Ashoka, and start attacking some of the most impactful problems of our age.
(Prof. Subhashis Banerjee was formerly the HoD for Computer Science. He plans to join back Computer Science department at Ashoka University in September 2023. Prof. Debayan Gupta is an Assistant Professor of Computer Science at Ashoka University)