India’s AI Mission: Ambitious Investments Amidst Accountability Concerns
The Indian government has recently approved a substantial budget of ₹10,371.92 crore to bolster its Artificial Intelligence (AI) infrastructure. While this move signifies a significant leap towards integrating AI into various state mechanisms, it also highlights a critical gap: the lack of compensation for individuals adversely affected by AI errors.
Case Study: The Plight of Dhuli Chand
A glaring example of AI-related administrative missteps is the case of Dhuli Chand, a centenarian from Gandhra village, Haryana. In March 2022, Chand’s old-age pension was abruptly halted after the Haryana Parivar Pehchan Patra (PPP) database erroneously marked him as deceased. In a bid to prove his existence, Chand organized a public procession in Rohtak city, showcasing documents such as his Aadhaar card and bank statements, demanding the restoration of his pension. Despite this public display, the error in the PPP database that led to this wrongful suspension remains unexplained, reflecting the opaque nature of AI systems, often termed as ‘black boxes.’
AI Guidelines: Accountability Without Compensation
In response to the challenges posed by the deployment of AI, the Indian government has introduced the India AI Governance Guidelines. These guidelines emphasize principles such as accountability and transparency for AI developers and users. However, they fall short in addressing the critical issue of compensation for individuals like Chand who suffer due to AI-induced errors. While the guidelines stress accountability, they don’t specify how the state should rectify or compensate for these mistakes.
Historical Context and Legal Challenges
The Aadhaar Act of 2016, which necessitated Aadhaar linkage for government subsidies, has had its share of controversies. There have been several instances where technological errors led to severe consequences, such as the tragic case of 11-year-old Santoshi Kumari in Jharkhand, who reportedly died of starvation after her family’s ration card was invalidated due to a missed Aadhaar linkage.
The Indian legal framework currently lacks specific provisions for attributing liability and ensuring compensation for damages caused by government AI systems. Even landmark judgments like Justice KS Puttaswamy v. Union of India, which solidified privacy rights, do not address compensatory mechanisms for AI errors.
International Comparisons and Lessons
Countries like Kenya and the European Union offer valuable lessons. The Kenyan High Court halted the Huduma Namba project until adequate legal safeguards were in place. Similarly, the EU AI Act classifies certain AI systems as high-risk and imposes substantial penalties for non-compliance, ensuring that liabilities are predefined and actionable.
These international frameworks underscore how preemptive legal provisions can protect citizens from the adverse effects of technological errors, a step that India has yet to take.
The Way Forward for India’s AI Framework
India’s investment in AI is undoubtedly a step forward, but it must be accompanied by a robust legal framework that addresses accountability and compensation. Without such measures, the deployment of AI will continue to pose risks to citizens, especially the most vulnerable, who may be unaware of their rights to rectification and compensation.
As India continues to advance its AI capabilities, it is imperative for the Parliament to establish statutory guidelines that allocate liability and ensure compensation, similar to the EU’s approach. This will not only address current deficiencies but also ensure that AI-driven initiatives are both innovative and responsible.
Nandini Agrawal is a student at UPES, Dehradun.
