India’s chief justice backs AI tools but rejects replacing judicial conscience

The Indian judiciary’s stance encapsulated here is likely to be influential well beyond India: AI can help judges think, but must not be seen as thinking for them. By explicitly framing AI as a support layer for research, translation and citizen interfaces, and pairing that with a proposed apex body for oversight, the CJI is sketching out a governance model where courts remain the human locus of legitimacy. Given India’s scale and its role as a reference for other common law jurisdictions, this could become a de facto template for “AI in justice” frameworks. ([dailyexcelsior.com](https://www.dailyexcelsior.com/ai-can-augment-reasoning-but-cannot-substitute-judicial-conscience-cji-surya-kant/?utm_source=openai))

For the AGI race, this kind of institutional stance is a reminder that not every domain will accept fully autonomous decision-makers, even if technically feasible. In practice, however, high-volume, low-stakes judicial tasks may slowly be delegated further to AI under such oversight bodies. That creates demand for specialised legal reasoning systems and benchmarks, as well as guardrails against subtle biases or hallucinations in legal contexts.