New Delhi: Telecom equipment suppliers selling into India may soon have to say where artificial intelligence sits inside their products. Draft guidelines issued by the Telecom Regulatory Authority of India on 30 October would make that disclosure part of the rules.
Under the draft, licensed operators must govern AI across their networks and services, whether they built the system or bought it. That means board-approved policies, a risk classification for each system, and a route for reporting serious AI incidents to the regulator.
TRAI said the draft carries the India AI Governance Guidelines, released by the government in February 2026, into the telecom sector. AI is being built into networks and services and can improve efficiency, the regulator said.
It can also bite. TRAI listed risks including bias introduced during training, model drift and adversarial attacks that could weaken the resilience of critical infrastructure. Putting a named owner on each system is the regulator's answer.
Written comments from the industry are open until 27 November, after which the final guidelines are expected.
The telecom industry has been moving fast on AI. Operators are using it for network planning, fault prediction and customer service. What has been less clear is accountability when a model gets something wrong.
For subscribers, the practical effect is narrower than it sounds but real. If an AI system decides which plan a customer is offered, how a complaint is routed or which network slice a call travels on, the operator will need a record of that system and a person answerable for it.