CJI Surya Kant warns of AI bias against underprivileged

The Intersection of Ancient Wisdom and Modern Jurisprudence: Justice Surya Kant on AI

In a profound discourse that bridged the gap between classical Indian literature and the cutting edge of modern technology, Justice Surya Kant of the Supreme Court of India recently delivered the 8th Dinkar Memorial Lecture. The lecture, centered on the timeless epic ‘Rashmirathi’ by the legendary poet Ramdhari Singh ‘Dinkar’, served as a platform for a critical warning: the rapid ascent of Artificial Intelligence (AI) poses a significant threat to the underprivileged due to inherent algorithmic biases. As a Senior Advocate observing the evolution of our legal landscape, it is clear that Justice Surya Kant’s observations touch upon the very soul of our Constitutional mandate—social justice.

The choice of ‘Rashmirathi’ as the thematic backdrop is particularly poignant. Dinkar’s magnum opus chronicles the life of Karna, a character who symbolizes the struggle against systemic exclusion and social hierarchy. By invoking this epic, Justice Surya Kant highlighted that while the tools of oppression may have changed from caste-based exclusion to algorithmic exclusion, the fundamental struggle for equity remains the same. The “digital divide” is not merely a matter of access to hardware; it is a matter of how software, specifically AI, perceives and treats those on the margins of society.

The Paradox of Progress: Efficiency vs. Equity

The core of the Justice’s warning lies in the paradox of technological progress. While AI is celebrated for its ability to process vast amounts of data and enhance administrative efficiency, it often operates within a “black box” that masks its prejudices. Justice Surya Kant cautioned that the growth of AI, if left unchecked by the principles of social justice, could inadvertently become a tool for further marginalizing the poor. This is not a hypothetical concern but a pressing reality in a world where AI influences everything from credit scoring and job recruitment to predictive policing and the distribution of welfare benefits.

As legal practitioners, we must recognize that AI systems are only as unbiased as the data they are trained on. In a society where historical and socio-economic disparities are deeply entrenched, the data generated by such a society is inevitably skewed. If an AI system is fed data that reflects systemic poverty or historical prejudice, it will learn to replicate and even amplify those biases. The result is a cycle of exclusion where the underprivileged are penalized not for their actions, but for a data-driven profile that views them as “high risk” or “low value.”

Algorithmic Bias and the Underprivileged

The “inherent bias” mentioned by Justice Surya Kant refers to the ways in which AI models can discriminate against individuals based on factors that correlate with poverty. For example, in automated systems used for determining eligibility for government schemes, an algorithm might prioritize certain postal codes or educational backgrounds, effectively filtering out those from rural or impoverished areas. This creates a digital glass ceiling that is much harder to break than traditional barriers because it is often invisible and defended as “objective” or “data-driven.”

Furthermore, the underprivileged often lack the “digital footprint” required to be accurately represented in AI training sets. This leads to what researchers call “data poverty.” When a segment of the population is underrepresented in the data, the AI fails to understand their specific needs and contexts, leading to inaccurate and often detrimental outcomes. This lack of representation is a modern form of the disenfranchisement that Dinkar’s Karna faced on the battlefield of Kurukshetra.

Rashmirathi as a Metaphor for the Digital Age

Justice Surya Kant’s invocation of ‘Rashmirathi’ serves as a reminder that the pursuit of justice is a perennial battle. In the epic, Karna is denied his rightful place due to his social standing, despite his incomparable merit. Similarly, in the age of AI, the underprivileged may find their merits overlooked by automated systems that prioritize data points over human dignity. The Justice emphasized that social justice continues to be the bedrock of a humane and equitable society, a principle that must not be sacrificed at the altar of technological advancement.

The epic poem asks difficult questions about identity, belonging, and the fairness of rules set by the powerful. By drawing these parallels, the Justice is urging the legal fraternity and policy-makers to look beyond the “hype” of AI and scrutinize its impact on the ground. Social justice is not a static concept; it must evolve to address the new forms of inequality generated by the Fourth Industrial Revolution.

The Role of the Judiciary in Policing AI

Under our Constitutional framework, the judiciary acts as the guardian of fundamental rights. Articles 14 (Equality before Law), 15 (Prohibition of Discrimination), and 21 (Protection of Life and Liberty) of the Indian Constitution are the pillars upon which we must evaluate the implementation of AI. Justice Surya Kant’s lecture suggests that the judiciary will not be a silent spectator to the digital marginalization of the poor. If an algorithm results in discriminatory outcomes, it must be held to the same standard of scrutiny as a discriminatory law or administrative action.

The challenge, however, lies in the complexity of AI systems. Traditional legal concepts of “intent” and “causation” are difficult to apply when a decision is made by a complex neural network. This necessitates a shift toward “algorithmic accountability.” We must demand transparency from those who deploy these systems, ensuring that the logic behind AI-driven decisions can be audited and challenged in a court of law. The bedrock of social justice requires that no person be condemned by a machine they cannot understand and an process they cannot appeal.

The Socio-Economic Implications of the Digital Divide

The “bias against the poor” is exacerbated by the widening digital divide in India. While urban centers enjoy high-speed connectivity and digital literacy, vast swathes of rural India remain underserved. When essential services—from healthcare to education—move to AI-driven platforms, those without access are not just left behind; they are effectively erased from the system. Justice Surya Kant’s warning highlights the danger of creating a “two-tier” society where technology empowers the elite while excluding the underprivileged.

In the context of the Dinkar Memorial Lecture, this divide is seen as a modern manifestation of the social barriers that Dinkar so passionately wrote against. The Justice’s speech is a call to action for the state to ensure that the benefits of AI are distributed equitably. This involves not only bridging the infrastructure gap but also ensuring that AI models are designed with the diverse socio-economic reality of India in mind.

AI in the Legal System: A Double-Edged Sword

The legal system itself is increasingly adopting AI for case management, research, and even predicting judicial outcomes. While this can help reduce the massive backlog of cases in Indian courts, Justice Surya Kant’s caution is particularly relevant here. If AI tools used by the judiciary or law enforcement carry inherent biases, it could lead to a systemic failure of justice. For instance, if predictive policing tools focus disproportionately on poor neighborhoods, they will inevitably result in higher arrest rates in those areas, which then feeds back into the AI as “proof” that those areas are high-crime, creating a self-fulfilling prophecy.

To prevent this, the legal profession must lead the way in establishing ethical guidelines for AI. We must ensure that AI is used to assist, not replace, human judgment. The “humane” element of justice, which Justice Surya Kant stressed, involves empathy, context, and an understanding of the human condition—qualities that an algorithm, no matter how sophisticated, cannot possess.

Constitutional Morality vs. Algorithmic Efficiency

The concept of “Constitutional Morality” is central to the Indian judiciary’s approach to governance. It dictates that the spirit of the Constitution must guide every action of the state. When Justice Surya Kant speaks of social justice as the “bedrock,” he is reaffirming that constitutional morality must take precedence over technological efficiency. An AI system that processes applications in seconds but discriminates against 20% of the population is a failure in the eyes of the law.

The underprivileged are often the most vulnerable to state action. When the state delegates its decision-making power to AI, it cannot delegate its constitutional responsibility. The Senior Advocate’s perspective is that we are entering an era where “due process” must include “digital due process.” This means that every individual has the right to an explanation of how an AI reached a decision affecting them, and the right to human intervention if that decision is biased.

Building an Inclusive AI Framework

To address the concerns raised by Justice Surya Kant, India needs a robust regulatory framework for AI that is rooted in social justice. This framework should include:

1. Mandatory Bias Audits: Any AI system used in public service or essential sectors must undergo regular audits to ensure it does not discriminate against the poor or other marginalized groups.

2. Diverse Training Data: Developers must be incentivized to use datasets that accurately represent the socio-economic diversity of India, ensuring that the underprivileged are not “invisible” to the algorithm.

3. Right to Redress: There must be clear legal pathways for individuals to challenge AI-driven decisions. This includes the establishment of specialized tribunals or ombudsmen who understand both technology and social justice.

4. Ethical Design: AI should be designed with “human-in-the-loop” systems, ensuring that final decisions, especially those affecting fundamental rights, are made by accountable human officials.

Conclusion: Upholding the Human Element

Justice Surya Kant’s lecture at the 8th Dinkar Memorial Event is a landmark moment in the Indian legal discourse on technology. By linking the struggle of Karna in ‘Rashmirathi’ to the modern challenges of AI bias, he has reminded us that the soul of justice is inherently human. Technology is a tool, but it is a tool that reflects the hands of its makers and the data of its history. If those hands are indifferent to the poor, and that history is one of inequality, the tool will inevitably become a weapon of exclusion.

As we move forward into an increasingly automated world, the legal fraternity must remain vigilant. We must ensure that the “bedrock of a humane and equitable society” is not eroded by the waves of technological change. Justice Surya Kant has set the tone for a necessary conversation: one where we embrace the future without forgetting the lessons of our past, and where we ensure that the light of progress reaches the last person in the queue, just as Dinkar’s poetry sought to give voice to the voiceless. The fight for social justice has merely moved to a new frontier—the digital one—and the principles of our Constitution remain our most reliable guide.

Ultimately, the warning is clear: AI must serve humanity, not the other way around. And “humanity” includes the millions of underprivileged citizens whose rights and dignity are just as sacred as those of the most technologically advanced elite. In the spirit of ‘Rashmirathi’, we must strive for a world where merit and dignity are not obscured by the shadows of bias, whether those shadows are cast by ancient prejudices or modern algorithms.