TRAI has proposed a new AI governance framework that could change how telecom operators use artificial intelligence in network management, customer services, fraud detection and other telecom operations. The draft guidelines require covered telecom entities to classify AI systems by risk, conduct impact assessments, maintain human oversight and report serious AI-related incidents to the regulator.
The Telecom Regulatory Authority of India (TRAI) released its draft Guidelines on Responsible Use of Artificial Intelligence (AI) in Telecommunication Networks and Provision of Telecommunication Services on October 30, 2026. The regulator has invited comments from stakeholders until November 27, 2026.
The proposal is particularly relevant to telecom operators such as Reliance Jio, Bharti Airtel and Vodafone Idea, which use AI and automation across network operations, customer support and digital services.
However, these are proposed guidelines, not final regulations. The requirements could change after TRAI reviews feedback from telecom companies and other stakeholders.
TRAI’s Proposed AI Governance Rules: Key Requirements
The draft framework proposes several responsibilities for telecom entities using AI in their networks or while providing telecom services.
| Proposed requirement | What telecom operators would need to do |
|---|---|
| AI risk classification | Classify AI systems as high-risk, medium-risk or low-risk |
| Internal AI policy | Create a board-approved policy governing the use of AI |
| Human oversight | Ensure appropriate human supervision of deployed AI systems |
| Impact assessments | Assess AI systems before deployment or modification |
| AI system inventory | Maintain records of active, inactive and decommissioned AI systems |
| Incident reporting | Record and investigate AI incidents and report critical and high-severity incidents to TRAI |
| Governance committee | Establish an AI Governance and Safety Committee |
| Consumer protection | Maintain a mechanism to address harm caused to telecom users by AI |
| Vendor accountability | Conduct due diligence and include regulatory safeguards in supplier agreements |
The proposed framework is intended to make telecom companies accountable for the way AI systems operate, irrespective of whether the technology is developed internally or purchased from an external vendor.
AI Systems Will Be Classified Into Three Risk Categories
One of the most important proposals is the classification of AI systems according to their potential impact on telecom networks and services.
TRAI wants covered entities to categorise each AI system as:
- High-risk: AI systems with significant implications for network operations, financial outcomes or consumers.
- Medium-risk: AI systems with a moderate level of operational or consumer impact.
- Low-risk: AI systems with relatively limited implications for telecom operations and users.
The classification would depend on the materiality of each system rather than simply the technology being used.
For example, an AI system involved in critical network operations could have a higher risk classification than an automated tool used for routine internal administrative tasks. Similarly, an AI system that influences customer access to telecom services or detects potentially fraudulent activity could require stronger oversight depending on its impact.
The proposed framework also requires companies to obtain approval from a designated officer before using an AI system. Approval responsibilities may be delegated to suitable officers depending on the system’s risk category.
Telecom Operators Must Assess AI Systems Before Deployment
TRAI’s draft proposes that telecom entities conduct validation, conformity assessments and impact assessments before deploying a new AI system or modifying an existing one.
These assessments would cover several parameters, including:
- Accuracy and reliability of the AI system.
- Robustness under different operating conditions.
- Explainability of its decisions and outputs.
- Fairness and the possibility of undesirable outcomes.
- Resilience against changes in model behaviour over time.
- Availability of appropriate human oversight.
This requirement could affect how telecom operators introduce AI-based network management systems, automated customer support tools and fraud detection platforms.
For example, if an AI system is used to identify suspicious activity or flag customer accounts, the operator would need to assess whether the system produces reliable results and whether its decisions could unfairly affect genuine customers.
The draft also proposes that telecom companies maintain an updated inventory of all AI systems, including those that are inactive or decommissioned. An AI system should not be used unless it is included in this inventory.
Jio, Airtel and Vi Could Face Greater Accountability for AI Decisions
The proposed guidelines make the telecom entity accountable for the outcomes of AI systems used in its operations.
This is an important provision because telecom companies increasingly depend on technology developed by external vendors and third-party service providers.
Under the draft, operators would need to conduct due diligence before acquiring or using third-party AI systems. This would include examining the credibility of suppliers, the limitations of their systems and the quality and suitability of training data.
The proposal also requires companies to include enforceable provisions in supplier agreements covering regulatory compliance, audit and inspection rights, incident management, risk allocation and accountability.
In practical terms, a telecom operator would not automatically escape responsibility for an AI-related failure simply because the technology was developed by an outside company.
AI Governance and Safety Committees Will Be Required
TRAI has proposed that covered telecom entities establish an AI Governance and Safety Committee (AGSC), with approval from their board of directors or an equivalent governing body.
The committee would include representatives from relevant business functions, including:
- Network operations.
- Service delivery.
- Information technology.
- Data management.
- Regulatory affairs.
- Consumer affairs.
The committee would be headed by a senior executive designated as the AI Governance and Safety Officer (AGSO).
Its responsibilities would include reviewing AI system performance, examining incident records, identifying vulnerabilities and assessing emerging risks. These reviews would take place at least once every financial year and whenever necessary.
If an AI system creates significant risks or repeatedly fails, the committee could recommend corrective measures, including modifying or decommissioning the system.
This would bring AI governance into the formal management and oversight structure of telecom companies rather than leaving it entirely to technology teams.
Critical AI Incidents Must Be Reported to TRAI
Another important part of the proposal concerns AI-related incidents.
TRAI wants telecom entities to classify incidents into four severity categories:
- Critical severity.
- High severity.
- Moderate severity.
- Low severity.
Companies would need to maintain a central database of AI incidents, investigate their causes and follow predefined procedures to mitigate the resulting harm.
Critical and high-severity incidents, including material malfunctions and recurring failures, would have to be reported to TRAI in the prescribed format and within the timeframe to be specified by the regulator.
The draft does not itself provide a fixed reporting deadline for these incidents. The regulator proposes to prescribe the reporting format and timeframe separately.
For telecom networks, this could become particularly important when AI-related failures affect network availability, service delivery, security or consumers.
The proposal also requires operators to maintain a separate record of AI-related harm reported by telecom users and how those complaints are resolved.
How Could the Proposed AI Rules Affect Telecom Customers?
Although the framework is primarily aimed at telecom operators, its implementation could have a direct impact on consumers.
AI is increasingly used in customer identification, fraud prevention, customer support, network optimisation and service management.
If the proposed guidelines are adopted, telecom companies would have to introduce stronger checks around how these systems are deployed and monitored.
For customers, the potential benefits include:
- More accountable automated decisions: Operators would remain responsible for AI systems even when they are supplied by third parties.
- Better monitoring of AI failures: Serious incidents would need to be recorded, investigated and reported.
- Human supervision: AI systems would be subject to appropriate human oversight.
- Consumer protection: Operators would need a defined mechanism to address AI-related harm.
- Improved system transparency: Risk assessments and governance procedures could help identify weaknesses before they affect customers.
However, the draft does not promise that every AI-powered customer service interaction will be handled by a human representative. Its focus is on appropriate oversight and accountability for AI systems used in telecom operations.
When Will TRAI’s AI Governance Guidelines Come Into Effect?
TRAI released the draft guidelines on October 30, 2026, and has invited stakeholder comments until November 27, 2026.
Stakeholders can submit their written comments electronically to the regulator. TRAI will consider the responses before deciding on the final framework.
The proposal should not be treated as an already implemented compliance requirement. The final guidelines could include changes based on feedback from telecom operators, technology companies and other stakeholders.
What the Proposed Rules Mean for India’s Telecom Industry
TRAI’s draft AI governance framework signals a move towards greater accountability as artificial intelligence becomes more deeply integrated into India’s telecom infrastructure.
For Jio, Airtel, Vi and other covered telecom entities, the proposed requirements could mean formal AI policies, risk assessments, governance committees, vendor checks and incident-reporting systems.
The biggest change is that telecom operators would remain accountable for the outcomes of AI systems even when the technology is developed or operated by third parties.
For consumers, the proposal could provide stronger safeguards against AI-related failures and unfair outcomes. However, the actual impact will depend on the final guidelines and how effectively the telecom industry implements them.
With the consultation deadline set for November 27, the next important development will be TRAI’s decision after reviewing stakeholder feedback.
