TRAI AI Guidelines 2026 - Proposed Compliance Requirements for Telecom Companies
Quick note
Below is the official summary and the reference document preview. Use βOpen PDFβ for full screen view.
Schedule a Call Back
Summary
The Telecom Regulatory Authority of India (TRAI) published the draft Guidelines on the Responsible Use of Artificial Intelligence in Telecommunication Networks and Provision of Telecommunication Services on 30 October 2026.
TRAI's AI Guidelines 2026 are designed to increase the accountability and safety of the use of artificial intelligence in telecommunication networks and services. They cover AI risk assessment, AI governance policy, AI systems assessment, AI incident reporting and protection of consumers.
Comments from stakeholders have been invited by TRAI till 27 November 2026.
Key Highlights of TRAI AI Guidelines 2026
The draft proposes several governance measures for authorised telecom entities using or planning to use AI systems.
- Internal AI Governance Policy: Covered entities would prepare an internal AI policy approved by their board or equivalent governing body.
- AI Risk Classification: AI systems would be classified as high-risk, medium-risk or low-risk.
- Deployment Approval: A designated officer would approve AI systems before their use.
- AI Systems Inventory: Companies would maintain records of active, inactive and decommissioned AI systems.
- Pre-deployment Assessments: New or modified AI systems would undergo validation, conformity assessment and impact assessment.
- AI Governance and Safety Committee: Covered businesses would establish a committee headed by a senior AI Governance and Safety Officer.
- AI Incident Reporting: Critical and high-severity incidents would be reported to TRAI.
- Third-Party Accountability: Firms would still be responsible for AI that was purchased or subcontracted to a vendor.
- Consumer Protection: Firms would put in place processes for dealing with consumer injury caused by AI.
- Annual Reviews: AI governance policies, risk classifications and committee reviews would take place at least once every financial year.
These measures are proposals under consultation, not requirements that have already taken effect.
What Are the TRAI AI Guidelines 2026?
The TRAI AI Guidelines 2026 are proposed sector-specific guidelines for the responsible use of artificial intelligence in India's telecommunications sector.
Telecom companies increasingly rely on AI for network optimisation, traffic management, fraud detection, customer support and operational planning. These applications can improve efficiency, but they can also create risks involving system reliability, data security, consumer rights and accountability.
For example, an AI system involved in network management may produce an incorrect recommendation or behave unexpectedly when operating conditions change. Depending on its function, such problems could affect network performance or service quality.
TRAI's proposed framework focuses on identifying these risks before deployment and managing them throughout the AI system's lifecycle.
Regulatory Framework Behind the Guidelines
The draft draws upon existing telecom legislation and India's broader AI governance framework.
| Law or guideline | Relevance |
| TRAI Act, 1997 | Establishes TRAI's regulatory responsibilities, including quality-of-service functions. |
| Indian Telegraph Act, 1885 | Relevant to telecom licences granted under the earlier framework. |
| Telecommunications Act, 2023 | Provides the newer telecom authorisation framework referred to in the draft. |
| India AI Governance Guidelines, February 2026 | Establish national AI governance principles referenced by TRAI. |
| TRAI Draft AI Guidelines, October 2026 | Propose sector-specific controls for AI systems used in telecom networks and services |
The draft refers to Section 11(1)(b)(v) of the TRAI Act, 1997.
It is important to understand that TRAI has not introduced a new universal AI licence or certification requirement through this consultation document.
Who Will Be Covered Under the TRAI AI Guidelines?
The proposed guidelines apply to entities holding specified telecom licences or authorisations that use or intend to use AI systems in telecommunication networks, telecommunication services or both.
The draft covers entities holding licences under the Indian Telegraph Act, 1885, or relevant authorisations under the Telecommunications Act, 2023.
Impact on Different Businesses
| Business category | Potential applicability |
| Telecom service providers | Covered where the specified licence or authorisation and AI-use conditions are met. |
| Telecom network operators | Relevant to authorised operators using or intending to use AI. |
| Internet service providers | May be covered depending on their licence or authorisation and AI use. |
| Telecom infrastructure businesses | Depends on the authorisation held and relevant AI operations. |
| AI software developers | Not automatically directly covered merely for supplying AI software. |
| Telecom equipment manufacturers | May face additional customer documentation and contractual expectations. |
Startups and MSMEs Direct coverage depends on their telecom authorisation and AI use.
A technology company supplying AI software to telecom operators does not automatically become directly regulated under these proposed guidelines.
However, telecom clients may require additional technical information, risk documentation and contractual assurances from their suppliers.
Major Compliance Requirements Proposed Under TRAI AI Guidelines 2026
1. Internal AI Governance Policy
One of the central proposals is the introduction of a formal internal AI governance policy. Under Section 3.1, covered telecom entities would formulate and implement an internal policy approved by their board of directors or equivalent governing body.
The policy would address AI risk classification, deployment approval, incident management, human oversight, data governance and consumer protection. The proposal also provides for reviewing the policy at least once every financial year and filing the policy or subsequent amendments with TRAI within seven working days of their coming into effect.
For telecom businesses, this could require closer coordination between management, compliance professionals, network engineers and technology teams.
2. AI Risk Classification
The draft proposes classifying AI systems into three categories.
| Risk category | General meaning |
| High-risk | AI applications with potentially significant operational, financial or consumer consequences. |
| Medium-risk | Systems presenting moderate potential consequences. |
| Low-risk | Systems with comparatively limited potential consequences. |
The actual classification would depend on the system's purpose, role and potential consequences. For example, an AI application influencing essential network operations could require closer scrutiny than a tool performing routine administrative functions. These examples are illustrative. TRAI has not provided a universal list assigning every telecom AI application to a fixed category. Companies would also review their risk classifications periodically.
3. Approval Before AI Deployment
The proposed framework requires approval from a designated officer before an AI system is used. Approval responsibilities may be delegated according to the risk category of the system.
This means companies may need formal procedures showing who is authorised to approve particular AI applications. Businesses using multiple AI tools across different departments could benefit from establishing a central approval mechanism rather than allowing unrecorded deployment decisions.
4. AI Validation and Conformity Assessment
Section 3.5(c) proposes validation, conformity assessment and impact assessment before deploying new AI systems or modifying existing systems. The draft identifies several important assessment parameters.
| Assessment parameter | Purpose |
| Accuracy | Evaluate the reliability of AI outputs. |
| Robustness | Assess performance under different operating conditions. |
| Explainability | Understand important factors influencing AI decisions. |
| Fairness | Identify potentially harmful bias. |
| Drift resilience | Evaluate deterioration caused by changing data or conditions. |
| Human oversight | Ensure appropriate supervision and intervention. |
The draft refers to TEC Standard 57050:2023 concerning fairness assessment and rating of AI systems, while allowing applicable international standards where appropriate.
It does not establish that every AI system must obtain a separate TEC certificate.
Telecom companies may nevertheless require additional technical assessments and records to demonstrate that their AI systems have been adequately evaluated.
5. AI Systems Inventory
The draft proposes maintaining an inventory containing information about all active, inactive and decommissioned AI systems. The inventory would record details such as the system's source, purpose and risk classification.
It also proposes that no AI system should be used unless included in the inventory. This could be particularly relevant for large telecom operators that use different AI platforms across network management, customer services and internal operations. A central inventory can help businesses identify unapproved tools, outdated systems and gaps in technical documentation.
6. AI Governance and Safety Committee
Section 3.6 proposes the establishment of an AI Governance and Safety Committee (AGSC), approved by the board or equivalent governing body. The committee would be headed by a senior executive serving as the AI Governance and Safety Officer (AGSO).
Representatives from relevant departments, including network operations, information technology, regulatory affairs and consumer affairs, would participate in governance activities. The committee would review AI systems at least once every financial year. Its responsibilities would include examining incidents, vulnerabilities, fairness concerns and emerging risks, and recommending appropriate corrective measures.
7. AI Incident Reporting Requirements
The proposed guidelines introduce four AI incident categories.
| Incident severity | Proposed requirement |
| Critical | Record, investigate, mitigate and report to TRAI. |
| High | Record, investigate, mitigate and report to TRAI. |
| Moderate | Record, investigate and mitigate through internal procedures. |
| Low | Record, investigate and mitigate through internal procedures. |
The draft also proposes maintaining a central database of AI incidents, including incidents arising during testing. TEC Standard 57090:2025 is referenced for AI incident database taxonomy, with the option of using an equivalent taxonomy.
Important: The draft does not specify a universal 24-hour or 72-hour reporting deadline. The form, manner and timeframe for reporting critical and high-severity incidents are to be prescribed.
8. Third-Party AI Systems and Vendor Accountability
Many telecom companies obtain AI-enabled software and equipment from external vendors. Under Section 3.8, covered authorised entities would remain accountable for AI systems used by them, even when those systems are developed or supplied by third parties.
The draft proposes supplier due diligence and appropriate contractual safeguards covering accountability, regulatory compliance, audit rights and incident management. This could affect how telecom operators select technology vendors and review existing agreements.
9. Data Governance and Consumer Protection
The proposed framework also addresses the lawful and secure handling of data throughout the AI lifecycle.
Companies would need to account for applicable privacy, confidentiality, cybersecurity and data-protection obligations. The draft also proposes mechanisms to record and resolve AI-related consumer harm.
A central database would help maintain information about complaints and their resolution. Although the proposal refers to resolving consumer harm within a reasonable timeframe, it does not prescribe a universal fixed number of days.
Existing Practices vs Proposed TRAI AI Requirements
| Compliance area | Existing position | Proposed guidelines |
| AI governance policy | Depends on applicable laws and existing organisational controls. | Board-approved internal AI policy. |
| Risk classification | May be handled through existing risk management processes | formal high-, medium- and low-risk classification. |
| AI deployment | Internal approval procedures may differ | designated-officer approval. |
| AI inventory | Depends on existing organisational requirements | central AI systems inventory. |
| Incident reporting | Existing telecom and cybersecurity obligations may apply | AI-specific reporting for critical and high-severity incidents. |
| Governance oversight | May be distributed across departments | dedicated AI Governance and Safety Committee |
| Vendor accountability | Governed by applicable laws and contracts. | Express accountability and supplier due diligence. |
The comparison describes the additional structure proposed by TRAI. It does not mean that telecom companies currently have no legal obligations relating to technology, data protection or cybersecurity.
TRAI AI Guidelines 2026 Deadline and Implementation Timeline
| Date or timeframe | Regulatory development |
| February 2026 | Release of India AI Governance Guidelines, as referenced by TRAI. |
| 30 October 2026 | TRAI released its draft telecom AI guidelines. |
| 27 November 2026 | Deadline for stakeholder comments. |
| Final implementation date | Not specified. |
| Seven working days | Proposed period for filing an internal AI policy or amendment after it comes into effect. |
| At least once every financial year | Proposed policy, risk-classification and governance committee reviews. |
How to Submit Comments to TRAI
Stakeholders can submit their written comments by 27 November 2026.
Businesses can consult the official TRAI website for the draft guidelines and consultation information. The consultation deadline should not be confused with an implementation deadline for telecom companies.
Why Has TRAI Proposed AI Guidelines for Telecom Companies?
Telecom networks support essential communication, commercial transactions, digital services and emergency response. As artificial intelligence becomes more involved in network management and service delivery, failures or unexpected AI behaviour may have significant consequences.
The draft identifies concerns including model drift, unintended outcomes, bias and adversarial attacks. Through the proposed guidelines, TRAI seeks to encourage responsible AI adoption supported by accountability, risk assessment, monitoring and human oversight.
The approach also draws upon seven principles from India's AI Governance Guidelines: trust, people-first development, innovation, fairness, accountability, understandability, and safety and resilience. The objective is to support AI innovation while addressing risks associated with its use in critical digital infrastructure.
Impact of TRAI AI Guidelines 2026 on Businesses
Operational Impact
Telecom operators may need to establish clearer procedures for introducing, monitoring and reviewing AI systems. This could require coordination between technology, legal, compliance and operations departments. Companies using numerous AI applications may also need to improve their system inventories and internal documentation.
Financial Impact and Compliance Costs
The preparation for the suggested framework could require costs for policy-making, technical analysis, employee training, management issues, and vendor verification. The actual cost will depend on the number of AI applications, their complexity and the organisation's existing systems. TRAI has not specified a new AI registration fee, mandatory certification fee or standard compliance consultancy charge in the draft.
Legal and Contractual Impact
Companies may need to review contracts with AI developers and equipment suppliers. The proposal places particular importance on accountability for outsourced AI systems. Depending on the final framework, businesses could need stronger contractual provisions concerning technical documentation, incident support and audit access.
Impact on AI Startups and Technology Suppliers
AI startups supplying products to telecom operators may face greater scrutiny during procurement. Telecom customers could request performance assessments, technical records and information about potential system risks. These are potential commercial implications rather than new direct legal obligations for all technology startups.
How Telecom Companies Can Prepare for TRAI AI Compliance
Although the guidelines are still under consultation, businesses can begin evaluating their AI governance readiness. The following steps are recommended preparations rather than immediately enforceable requirements under the draft.
- Check Applicability: Review the company's telecom licences or authorisations and identify whether its AI activities fall within the proposed scope.
- Identify AI Systems: Prepare an inventory of AI tools used in network operations, service delivery and relevant internal functions.
- Conduct a Compliance Gap Assessment: Compare existing governance controls with the requirements proposed by TRAI.
- Develop an AI Governance Policy: Prepare appropriate documentation covering risk classification, approvals, incident management and oversight.
- Assess AI Risks: Review the potential consequences of AI failures and identify systems requiring detailed technical assessment.
- Review Third-Party Vendors: Evaluate vendor documentation and contractual obligations in relation to incident response obligations.
- Plan Governance Responsibilities: Consider how the governance and oversight of the new AGSC and AGSO will fit within the existing organizational structure.
- Plan Incident Response Procedures: Develop appropriate record-keeping, escalation procedures, and investigative procedures.
- Train Employees: Increase employee knowledge regarding the governance of artificial intelligence.
- Monitor the Final Guidelines: Update the compliance programme after TRAI issues its final decision and implementation instructions.
Documents Businesses May Need
| Document | Purpose |
| Internal AI governance policy | Establishes AI management and accountability procedures. |
| AI systems inventory | Records AI applications, functions and risk categories. |
| Risk assessment records | Supports identification and classification of AI-related risks. |
| Conformity assessment reports | Documents relevant technical assessments. |
| AI incident register | Records incidents, investigations and corrective actions. |
| Vendor due diligence records | Supports evaluation of third-party AI suppliers. |
| Consumer complaint records | Tracks AI-related consumer harm and resolution. |
| Governance committee records | Supports management oversight and periodic review. |
Some records are expressly contemplated by the draft, while others are supporting documents that may help businesses demonstrate good governance. TRAI has not prescribed a universal application-document checklist.
Benefits of Responsible AI Governance for Telecom Businesses
Preparing for AI governance can help businesses manage technology-related risks more systematically.
- Improved Network Reliability: Appropriate testing and monitoring can identify system weaknesses before they affect important operations.
- Better Regulatory Readiness: Organised documentation and governance procedures can make future compliance implementation easier.
- Stronger Accountability: Clearly assigned responsibilities help management oversee the use of AI systems.
- Improved Vendor Management: Supplier assessments and appropriate agreements can reduce uncertainty about outsourced technology.
- Consumer Confidence: Transparent complaint-handling procedures can improve how businesses respond to AI-related concerns.
These are potential organisational benefits, not guaranteed outcomes.
Are TRAI's AI Guidelines a Positive Step or an Additional Compliance Burden?
The proposed guidelines could strengthen accountability and consumer protection in the telecom sector. Formal risk assessments, incident records and governance responsibilities may help companies identify AI-related problems earlier and manage them more effectively. However, implementation could also require additional resources.
Smaller authorised telecom businesses may face difficulties arranging specialist assessments, governance personnel and documentation systems. The draft also leaves certain matters open, including detailed incident-reporting timeframes and precise risk classification methodologies. These issues may require further clarification during the consultation process. A proportionate final framework could help businesses manage AI risks without creating unnecessary barriers to technology adoption.
Business Opportunities Under the Proposed TRAI AI Framework
The draft may increase demand for AI governance documentation, regulatory advisory, technical risk assessments, supplier due diligence and compliance-readiness services. Technology providers offering well-documented AI systems may also find it easier to meet telecom customers' procurement expectations.
However, the draft does not create a separate compulsory AI consultancy licence or universal TRAI AI certification scheme. Businesses should evaluate these opportunities based on the final regulatory requirements and genuine operational needs.
How Corpseed Can Help with TRAI AI Compliance
Telecom companies may require professional support to understand the proposed guidelines and assess their existing governance arrangements.
Corpseed can assist businesses through relevant regulatory consultancy and compliance documentation support.
- Regulatory Compliance Consultancy: Guidance on the proposed TRAI AI framework and its applicability.
- AI Governance Documentation: Assistance with internal policy documentation and compliance procedures.
- Compliance Gap Assessment: Identifying differences between existing practices and proposed regulatory expectations.
- AI Risk Assessment Coordination: Helping organise documentation and coordinate relevant specialist evaluations.
- Vendor Compliance Support: Assistance with supplier-related compliance documentation.
- Regulatory Advisory: Guidance on relevant telecom regulatory requirements and emerging obligations.
- Ongoing Compliance Support: Monitoring regulatory developments and assisting with compliance-readiness documentation.
Professional guidance can help businesses distinguish mandatory legal requirements from proposed measures and prepare proportionate compliance arrangements.
Need help understanding how TRAI's proposed AI guidelines may affect your telecom business? Contact Corpseed for telecom regulatory compliance consultancy and AI governance documentation support.
Conclusion
The TRAI AI Guidelines 2026 represent an important proposal for strengthening artificial intelligence governance in India's telecommunications sector. The draft covers internal AI policies, risk classification, system assessments, incident management, vendor accountability and consumer protection.
For telecom companies, the immediate priority is to understand the proposals, evaluate their potential impact and consider providing comments before 27 November 2026. Since the guidelines have not been finalised, businesses should avoid treating proposed obligations as current mandatory requirements.
Corpseed can help affected organisations understand the regulatory framework, assess governance gaps and prepare relevant compliance documentation.
Schedule a Call Back
Document Preview
Embedded reference document
