The RBI panel has recommended a “tolerant supervisory stance” towards AI errors in financial systems to foster innovation while ensuring adequate safety measures are in place. The report emphasizes that AI’s probabilistic nature warrants some leeway for first-time mistakes, arguing that strict penalties could hinder progress. It outlines seven guiding principles, including maintaining trust and ensuring fairness, alongside 26 recommendations focusing on enhancing data infrastructure, creating testing sandboxes, and implementing robust governance and cybersecurity measures. This balanced approach aims to encourage responsible AI development without compromising consumer protection.