TRAI Drafts AI Rules for Telecom Networks
The regulator proposes algorithmic transparency disclosures and a mandatory independent AI audit regime for operators.
The Telecom Regulatory Authority of India (TRAI) has released a consultation paper and draft framework for using artificial intelligence in telecom networks — the first sector-specific AI governance proposal by an Indian regulator.
It covers AI used for network optimisation, customer experience, fraud detection, content moderation and service-bundling pricing.
Key Highlights
- India's first sector-specific AI governance framework.
- Mandatory algorithmic transparency disclosures to TRAI.
- Independent AI audits on a two-year cycle via an empanelled registry.
- Penalties for AI systems with discriminatory service outcomes.
- Audits check bias, data security and DPDP Act compliance.
Algorithmic Transparency
The core proposal requires licensed operators deploying AI in network management or subscriber-facing functions to register each model's purpose, training-data sources, decision criteria and accuracy metrics with TRAI via a standard template, updated annually or on material change.
Independent Audit Regime
The framework proposes a TRAI-empanelled AI Audit Registry, requiring operators to commission independent technical audits every two years. Audits examine algorithmic bias — especially bias that disadvantages rural or low-income subscribers — plus data security and Digital Personal Data Protection Act compliance.
Operators running AI with demonstrable discriminatory outcomes would face penalties scaled to subscriber base and severity. The paper cites cases abroad where network-slicing algorithms deprioritised emergency traffic in lower-income areas as the harm it aims to prevent.
What the framework proposes
| Element | Requirement |
|---|---|
| Transparency | Register model purpose, data, criteria, accuracy |
| Audits | Independent, every two years |
| Focus | Bias, data security, DPDP compliance |
| Penalties | Scaled to subscriber base and severity |
Industry Response
The Cellular Operators Association of India welcomed the intent but flagged compliance costs for smaller operators and questioned TRAI's in-house capacity to evaluate audit reports, suggesting partnership with the IITs on standardised audit methods.
Digital-rights groups backed the proposal and urged TRAI to extend it to network-level content decisions, which they said carry free-expression implications. The move parallels India's broader tech-governance agenda alongside its 6G roadmap.
Frequently Asked Questions
What does the TRAI framework cover?
AI used in telecom for network management, customer service, fraud detection, content moderation and pricing.
What is algorithmic transparency here?
Operators must register each AI model's purpose, data, decision criteria and accuracy with TRAI.
How often are audits required?
Every two years, by independent auditors on a TRAI-empanelled registry.
What happens for biased AI?
Financial penalties scaled to subscriber base and the severity of demonstrated bias.
Is this final?
No. It is a draft framework open for consultation.
Sources
- Telecom Regulatory Authority of India — consultation paper
- Cellular Operators Association of India