Supreme Court judges are highly vigilant about AI risks like "hallucinations" (fabricated precedents) and biases, using tools such as SUPACE and AI Saransh only as aids for research, transcription, and analytics, with mandatory human verification to prevent over-reliance.
The bench dismissed a PIL for regulatory guidelines, deeming concerns valid but best addressed administratively through judicial training and the court's White Paper, inviting public suggestions while stressing AI must never eclipse judicial reasoning or empathy.
Echoing precedents like the Kerala High Court's AI policy, the ruling reinforces e-Courts Phase III's tech integration for tackling India's 50M+ case backlog, but prioritizes data security, transparency, and the irreplaceable role of human judgment in upholding constitutional justice.
Supreme Court Jurisdiction On AI
Supreme Court vows vigilance against AI's "hallucinations" and biases, ensuring algorithms assist but never dictate justice
The Supreme Court of India on Friday declared that judges are acutely aware of AI's pitfalls and will never allow it to overshadow human reasoning in the dispensation of justice. Chief Justice of India Surya Kant, leading a bench alongside Justice Joymalya Bagchi, dismissed a public interest litigation (PIL) seeking mandatory guidelines to regulate AI's use in the judiciary, emphasizing that the court is already handling such concerns through administrative measures and rigorous training protocols.
CJI Kant, however, was unequivocal: "We use it in a very over-conscious manner and we do not want this to overpower our judicial decision-making." He underscored that while AI tools such as the Supreme Court Portal for Assistance in Courts Efficiency (SUPACE) and AI Saransh are invaluable for sifting through massive case backlogs – India's courts grapple with over 50 million pending matters – they remain strictly assistive. "Judges must cross-check. This is part of the judicial academy curriculum and is taken care of," the CJI added, noting that the Bar too must evolve to verify AI outputs, turning such errors into teachable moments.
Dismissing the PIL as not "judicially actionable," the bench permitted its withdrawal and extended an open invitation: "Someone with sincere intentions is most welcome to give us suggestions. You can mail them to us." This administrative pivot signals a proactive, self-regulatory approach, aligning with global concerns over AI in justice systems while prioritizing India's constitutional ethos of fair, human-centered adjudication.