Indian Supreme Court drafts strict rules for AI in courts

India’s top court is drawing a bright line between AI as an assistive tool and AI as a decision‑maker in the justice system. The draft rules effectively treat adjudication as a “high‑risk” application where human primacy must be preserved, even as courts lean on AI for transcription, translation and case management. That mirrors emerging global norms but from a jurisdiction with enormous case volumes and a rapidly digitising judiciary.

For the AGI race, this is a reminder that deployment constraints can matter as much as model capabilities. Even if future systems could in principle out‑reason human judges on some metrics, a major democracy is pre‑committing to keep them out of the decision loop. That doesn’t slow core frontier-model research, but it does narrow one lucrative application domain and signals that other high‑stakes sectors—criminal justice, credit scoring, social benefits—may see similar guardrails.

Strategically, India is also using this moment to shape professional norms: lawyers and lower courts are being told they remain fully responsible for checking AI outputs and citations after several cases involving hallucinated judgments. That will influence how Indian legal-tech startups position their tools and how global vendors design “court‑safe” versions of their models.