Auditing the Automated State: Impact Assessments and Institutional Capacity for Ethical AI in Indian Governance
Author(s): Lt. Dr. Kongala Sukumar
Publication #: 2608014
Date of Publication: 05.08.2023
Country: India
Pages: 1-13
Published In: Volume 9 Issue 4 August-2023
DOI: https://doi.org/10.62970/IJIRCT.v9.i4.2608014
Abstract
As Indian public administration increasingly delegates welfare targeting, policing, and service delivery to algorithmic systems, the absence of a statutory impact-assessment regime has left citizens with limited means of contesting automated decisions. This paper examines Algorithmic Impact Assessments (AIAs) as a governance instrument capable of anchoring ethical artificial intelligence within Indian public institutions, and asks whether existing administrative capacity can sustain such a regime. Drawing on a comparative analysis of the Canadian Directive on Automated Decision-Making, the European Union's data-protection impact-assessment architecture, and emergent Indian policy instruments issued by NITI Aayog, the paper develops a four-dimensional institutional capacity framework spanning technical expertise, legal authority, procedural infrastructure, and civic accountability. Three illustrative domains of automation in Indian governance — direct benefit transfer, predictive policing, and facial recognition — are examined against this framework. The analysis finds that India's AI policy discourse has matured considerably since 2018, yet implementation remains fragmented across ministries, with no binding pre-deployment review requirement in force. The paper concludes with a phased institutional roadmap for embedding impact assessments within India's administrative law architecture, emphasising capacity-building, sectoral regulatory coordination, and public registries of automated systems as preconditions for accountable algorithmic governance.
Keywords: algorithmic impact assessment; ethical artificial intelligence; institutional capacity; Indian public administration; administrative accountability; automated decision-making; NITI Aayog
Download/View Count: 3
Share this Article