Indian legal leaders call for human oversight in AI systems

A call for global cooperation

The 4th Law and Constitution Dialogue 2026 in New Delhi brought together India's top legal figures to discuss the rapid rise of artificial intelligence. Chief Justice of India Justice Surya Kant noted that judiciaries worldwide are currently exploring how to use AI for tasks like legal research, translation, and case management.

Justice Kant argued that because every country is learning about these technologies at the same time, it is inefficient for courts to experiment in isolation. He advocated for greater international cooperation to assess the long-term implications of these tools before they are adopted on a wider scale.

To facilitate this, he proposed a Judicial Innovation Lab. This platform would allow different judicial institutions to jointly examine emerging technologies and identify their specific capabilities and risks in a shared environment.

Preserving human wisdom

Union Law Minister Arjun Ram Meghwal addressed the shift toward 'agentic AI,' which refers to systems capable of taking more independent actions. He raised concerns about the possibility of AI suggesting specific arguments for legal parties or even predicting how a judge might rule on a case.

The minister insisted that technology should serve as an assistant to human beings rather than a replacement for them. He linked India's technological growth to its philosophical traditions, suggesting that progress should aim for what he called 'spiritual intelligence' alongside artificial intelligence.

Justice Kant echoed this sentiment by distinguishing between intelligence and wisdom. He stated that true justice is not served by intelligence alone, but by the human understanding and judgment required to administer the law fairly.

Addressing bias and privacy

Senior advocate Mahesh Jethmalani noted that tools like Claude and ChatGPT are already being integrated into the legal profession. While he cautioned against exaggerated fears, he warned that the rule of law must harness technology rather than being controlled by it.

Jethmalani highlighted specific risks including privacy violations, deepfakes, and the misuse of synthetically generated content. He pointed to India's Digital Personal Data Protection Act and upcoming 2026 rules as necessary frameworks to provide penalties and takedown mechanisms for these threats.

The discussion also touched on the danger of algorithmic bias. Experts warned that if an AI is trained on flawed historical data, it might present old injustices as objective predictions for the future, effectively automating past discrimination.

Legislating for the future

Sasmit Patra, a Member of Parliament, emphasized the need for lawmakers to develop a deeper technical understanding of algorithms before drafting new regulations. He noted a significant gap between the influence of AI and the level of technological knowledge currently held by many legislators.

Despite these concerns, Patra suggested that AI could actually strengthen the legislative process. It has the potential to help with research, identify inconsistencies in complex regulatory frameworks, and assess the potential consequences of new policies.

However, he maintained that any decisions involving public interest or citizens' rights must remain guided by human judgment. He reminded the assembly that technological progress must stay within constitutional boundaries, specifically protecting the rights guaranteed under Articles 14, 19, and 21.

The anonymity of responsibility

A central theme of the dialogue was the need for a comprehensive global legal framework to establish accountability for autonomous systems. Rajya Sabha MP Dr. Abhishek Manu Singhvi warned that increasing technological autonomy must not be used as a shield to escape responsibility for the consequences of AI-driven decisions.

He cautioned that if machines are allowed to operate without clear human oversight, it could lead to an 'anonymity of responsibility.' This means that when an automated system makes a mistake or causes harm, it must be clear who is held accountable for that outcome.

The participants concluded that while AI offers significant opportunities for efficiency, it must be governed by constitutional morality. The goal is to ensure that technology remains a servant of justice, ensuring that human reasoning remains at the heart of the legal system.