Saturday, July 25, 2026 Live
← Back to Home
Technology Verified · 2 sources

National AI safety institute opens, sets first model-audit rules

The new body will run voluntary evaluations of large AI systems before release and publish shared benchmarks for reliability and misuse risk.

BENGALURUBy Tech DeskJul 22, 2026

A government-backed AI safety institute began operations this week, tasked with evaluating large AI models for reliability and misuse risk before they are widely deployed.

Its first framework sets out voluntary audits and a shared set of public benchmarks. Officials said participation would be opt-in initially, with the aim of building industry norms rather than immediate mandates.

This is illustrative sample content included with the BreakNewsLine scaffold.

Verified against 2 independent sources
  1. 01

    Founding mandate and first audit framework.

  2. 02
#ai#regulation