New Delhi, October 31, 2026: The Telecom Regulatory Authority of India (TRAI) has proposed a risk-based governance framework for artificial intelligence used in telecom networks and services. The regulator released draft guidelines for consultation on October 30 and has invited stakeholder comments until November 27, 2026.
The proposal applies to entities holding a telecom licence or authorisation under the Telegraph Act, 1885, or the Telecommunications Act, 2023, that use or plan to use AI in telecom networks or services. It is not a general rule for every AI developer or software company.
Key requirements proposed by TRAI
- Board-approved AI policy: Applicable telecom entities would need an internal AI policy approved by their board or equivalent governing body.
- Risk classification: Each AI system would be classified as high, medium or low risk, based on its importance to network operations and its potential impact on consumers.
- Deployment approval and oversight: AI systems would require approval from a designated officer and appropriate human oversight.
- Incident reporting: Companies would maintain a central database of AI incidents. Critical- and high-severity incidents would have to be reported to TRAI in a prescribed format and timeframe.
- Vendor checks and impact assessments: Entities would maintain an inventory of active, inactive and retired AI systems, conduct due diligence on third-party suppliers, and assess accuracy, robustness, fairness and explainability before deployment or modification.
- Governance committee: A board-approved AI Governance and Safety Committee, headed by a senior executive, would oversee implementation and review performance at least once each financial year.
The draft also proposes data-governance safeguards, a mechanism to address AI-related consumer harm, and annual reviews of internal AI policies. TRAI says AI can improve telecom efficiency, but inadequate oversight may expose critical infrastructure to bias, model drift, adversarial attacks and other failures.
Comments open until November 27
These are draft guidelines for consultation, not final binding rules. Stakeholders can submit written comments by November 27, 2026, through TRAI’s consultation process or by email to advmn@trai.gov.in, as specified in the consultation paper.
Business Standard and Mint also reported the proposed board-level policies, risk classification and incident-reporting framework. Business Standard noted that the draft leaves some incident-severity thresholds and reporting timelines to be prescribed.
Sources
TRAI consultation page · Official draft consultation paper (PDF) · Business Standard · Mint.
