Indian Experts Call for Domestic AI Safety Framework

Industry experts say India should create an AI safety framework suited to its languages, scale and focus on AI applications, according to a PTI report published by ET BFSI. They cautioned against directly copying the voluntary White House Accord on Super Intelligence signed on September 29 by major US technology companies.

NASSCOM Chairman Srikanth Velamakanni called for binding requirements covering transparency, incident reporting and accountability for serious harm. Other experts supported a hybrid approach that would apply mandatory rules to uses in areas such as defense, health care, banking and government, while allowing voluntary standards for rapidly changing technical issues.

Suggested safeguards include testing before deployment, independent audits, named responsibility for AI risks and reliable performance across Indian languages and cultural contexts. Experts identified deepfakes, cyberattacks, fraud, privacy breaches and attacks on critical infrastructure as key risks, though some favored self regulation to avoid limiting competition.