India’s telecom regulator, the Telecom Regulatory Authority of India (TRAI), has proposed new rules that would make telecom operators accountable for failures involving artificial intelligence systems used in their networks and services. Under the draft guidelines released on October 30, 2026, companies such as Bharti Airtel, Reliance Jio and Vodafone Idea (Vi) would have to report critical and high-severity AI incidents to the regulator, maintain records of failures and establish formal systems for overseeing AI deployment.
The proposal is part of TRAI’s draft Guidelines on Responsible Use of Artificial Intelligence in Telecommunication Networks and Provision of Telecommunication Services. It also calls for board-approved AI policies, risk-based classification of AI systems, human supervision and assessments before new AI tools are deployed. Stakeholders have until November 27, 2026, to submit comments. The guidelines are proposals, not yet final regulations.
Key Takeaways
- TRAI released draft AI governance guidelines for the telecom sector on October 30, 2026.
- Airtel, Jio, Vodafone Idea and other covered telecom entities could be required to report critical and high-severity AI failures.
- Operators would remain accountable even when AI systems are developed or supplied by third-party vendors.
- Companies would need board-approved AI policies and dedicated AI governance committees.
- AI systems would be classified as high, medium or low risk.
- Operators would have to maintain inventories of their AI systems and records of incidents.
- Stakeholder comments are due by November 27, 2026.
What TRAI Is Proposing for Telecom Companies
TRAI wants telecom operators to establish formal governance systems for the artificial intelligence technologies they use.
The draft applies to entities holding licences under the Indian Telegraph Act, 1885, or authorisations under the Telecommunications Act, 2023, that use or intend to use AI in telecom networks and services.
The proposed framework covers AI systems used in areas such as network management, customer services, fraud detection, security and operational decision-making.
Under the proposal, each covered company would need an internal AI policy approved by its board of directors or equivalent governing body.
These policies would establish how AI systems are selected, assessed, deployed, monitored and discontinued.
The policies would also have to address human oversight, data governance, consumer protection and complaint resolution.
Serious AI Failures Must Be Reported to TRAI
One of the most important parts of the proposal concerns AI-related incidents.
TRAI wants telecom companies to classify failures into four categories:
| Incident category | Proposed treatment |
|---|---|
| Critical | Mandatory reporting to TRAI |
| High | Mandatory reporting to TRAI |
| Moderate | Record, investigate and manage |
| Low | Record and manage through internal procedures |
Critical and high-severity incidents, including material malfunctions and recurring failures, would have to be reported to TRAI in a prescribed format and within a prescribed timeframe.
The draft also requires companies to maintain a central database of AI incidents.
This database would help operators document failures, investigate their causes and implement corrective measures.
The reporting requirement is particularly important because AI systems increasingly influence how telecom networks operate. A malfunction could affect network availability, customer services, fraud detection or other critical functions.
However, the draft does not specify exact reporting deadlines or provide a complete list of events that automatically qualify as critical or high severity. These details may be clarified during the consultation process.
Airtel, Jio and Vodafone Idea Could Be Accountable for Vendor AI
Another significant provision concerns third-party AI systems.
Telecom companies frequently purchase network equipment, software and technology services from external vendors. Some of these products include AI capabilities that operators may not develop themselves.
TRAI’s proposal seeks to make the telecom entity accountable for the outcomes of AI systems used in its networks and services, regardless of whether those systems are developed internally, purchased from vendors or operated through a combination of both.
This means operators could not simply shift responsibility to a technology supplier when an AI system causes a serious problem.
Companies would also have to obtain information about AI systems embedded in equipment they purchase and include appropriate audit and accountability provisions in supplier contracts.
The proposal could therefore have implications for both telecom operators and the companies supplying their network infrastructure.
Companies Must Classify AI Systems by Risk
TRAI wants telecom operators to classify their AI systems according to their potential impact.
The proposed categories are:
High-risk AI: Systems with significant implications for network operations, service delivery, financial or operational outcomes, or consumers.
Medium-risk AI: Systems with a more limited but still meaningful operational or consumer impact.
Low-risk AI: Systems with comparatively limited consequences if they malfunction.
The classification would help determine the level of scrutiny and approval required before an AI system is deployed.
The proposal also requires companies to maintain updated inventories of their AI systems, including active, inactive and discontinued applications.
Each entry would document the system’s source, purpose and risk classification. An AI system would not be permitted for use unless it appeared in the company’s inventory.
Board-Level AI Governance and Human Supervision
TRAI wants AI accountability to become a formal management responsibility rather than an issue handled only by technical teams.
Under the draft, telecom entities would have to establish an AI Governance and Safety Committee.
The committee would include representatives from areas such as:
- Network operations
- Service delivery
- Information technology
- Data management
- Regulatory affairs
- Consumer affairs
A senior executive would lead the committee as the AI Governance and Safety Officer.
The proposed framework also requires appropriate human oversight of AI systems through predefined mechanisms.
This is intended to ensure that important decisions and potentially harmful outcomes can be monitored and addressed by people rather than being left entirely to automated systems.
AI Systems Must Be Tested Before Deployment
The draft guidelines also focus on preventing problems before AI systems enter production.
Before deploying a new AI system or making significant changes to an existing one, telecom entities would need to conduct validation, conformity and impact assessments.
These assessments would cover several parameters:
| Assessment area | What it examines |
|---|---|
| Accuracy | Whether the AI produces reliable results |
| Robustness | Whether the system performs consistently |
| Explainability | Whether its decisions can be understood |
| Fairness | Whether it produces biased or discriminatory outcomes |
| Model resilience | Whether performance remains reliable as conditions change |
Companies would also need procedures to investigate AI-related complaints and address harm reported by consumers within a reasonable timeframe.
These requirements are intended to reduce risks such as biased decisions, unexpected model behaviour, model drift and adversarial attacks.
How Airtel, Jio and Vodafone Idea Already Use AI
The proposal comes as Indian telecom operators increasingly integrate AI into their networks and customer services.
Bharti Airtel: The company uses AI-powered systems to identify spam calls and messages. It has also introduced tools designed to detect malicious links and protect users against digital fraud.
Reliance Jio: Jio uses its in-house AI platform, JioBrain, for network-related functions such as fault prediction, capacity allocation and network optimisation.
Vodafone Idea: Vi uses AI-based systems to identify spam communications and strengthen cybersecurity through its Vi Protect initiative.
These applications demonstrate why AI governance is becoming important for telecom companies. An error in a spam-detection system may inconvenience customers, while a malfunction affecting network operations or security could have wider consequences.
The proposed rules would introduce a more structured process for monitoring and reporting such risks.
What Happens Next?
TRAI has invited comments from stakeholders on the draft guidelines until November 27, 2026.
Companies, technology vendors, industry associations and other interested parties can submit their feedback to the regulator.
The regulator will consider these responses before deciding whether to modify or finalise the proposed framework.
The consultation process will be important for clarifying practical questions, including incident-reporting deadlines, risk-classification criteria and the responsibilities of telecom operators when third-party AI systems fail.
The Bigger Picture
TRAI’s proposal reflects a broader shift in how regulators are approaching artificial intelligence. Instead of focusing only on the companies that develop AI models, regulators are increasingly examining the organisations that deploy these systems in essential services.
Telecommunications is particularly important because mobile and broadband networks support banking, emergency communication, commerce, government services and everyday digital activity.
If an AI system disrupts network operations or incorrectly handles an important customer function, the consequences can extend beyond the technology provider.
By proposing board-level oversight, incident reporting and operator accountability, TRAI is attempting to establish clearer responsibility for AI-related risks in India’s telecom infrastructure.
Looking Ahead
The November 27 consultation deadline is the next important milestone. Airtel, Jio, Vodafone Idea, equipment suppliers and industry groups may seek clarification or changes to the proposed requirements. The final guidelines could differ from the current draft, particularly around reporting deadlines, compliance procedures and the classification of serious incidents.
If adopted, the framework would make AI governance a more formal part of telecom operations in India. Operators would need stronger systems for monitoring AI performance, investigating failures and documenting corrective action. For consumers, the intended benefit is greater accountability when automated systems affect network services, security or customer experience.
Frequently Asked Questions
1. What has TRAI proposed for Airtel, Jio and Vodafone Idea?
TRAI has proposed AI governance guidelines requiring telecom operators to classify AI systems by risk, maintain incident records, provide human oversight and report critical and high-severity AI failures to the regulator.
2. When did TRAI release the draft AI guidelines?
TRAI released the draft Guidelines on Responsible Use of Artificial Intelligence in Telecommunication Networks and Provision of Telecommunication Services on October 30, 2026.
3. Which AI failures must telecom companies report?
The proposal requires reporting of critical- and high-severity incidents, including material malfunctions and recurring failures. The exact reporting timeframe and detailed classification criteria remain to be clarified.
4. Will telecom companies be responsible for AI developed by vendors?
Yes. The draft proposes that telecom entities remain accountable for AI systems used in their networks and services, whether those systems are developed internally or supplied by third parties.
5. What is an AI Governance and Safety Committee?
It is a proposed internal committee responsible for overseeing AI-related risks, compliance and safety. It would include representatives from technical, operational, regulatory and consumer-facing departments.
6. Will the proposed TRAI rules apply immediately?
No. These are draft guidelines undergoing stakeholder consultation. They are not final regulations yet.
7. What is the deadline for submitting comments?
TRAI has invited stakeholder comments until November 27, 2026.
8. How could the proposal benefit telecom customers?
The proposed framework could improve accountability when AI systems cause service disruptions, security problems or consumer harm. It also seeks to establish clearer procedures for investigating and resolving AI-related complaints.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.
