TRAI makes telcos accountable for outsourced AI failures
Telecom operators, not vendors, would carry the main regulatory burden
Overview
Telecom operators could soon bear the legal and financial risk for every AI system they use, even when a vendor built it. released on October 30, 2026, and will accept comments until November 27. The rules are not final, but they would require board-approved AI policies, oversight committees, , tests for accuracy and fairness, human review, complaint systems, and about serious failures.
Operators would also need stronger supplier contracts covering audits, responses to problems, and who pays when systems fail. This could improve safety and give customers clearer ways to seek help, but it could slow new AI services and raise costs. Large operators may manage better because they have bigger compliance teams and more bargaining power. Smaller operators and vendors could face heavier burdens. The rules for reporting failures are still open to public comment.
Timeline
TRAI's proposed rules would add AI governance and reporting duties for telcos
If adopted, the draft framework would require telecom entities to maintain AI policies and inventories, assess and monitor systems, investigate incidents, and report serious failures. It would also require due diligence and accountability provisions for .
TRAI set November 27 as the deadline for stakeholder comments on telecom AI rules
Stakeholders must submit written comments on the draft guidelines by November 27, 2026. The will test how the proposed , reporting duties, vendor obligations and governance controls can be applied across operators and suppliers.
TRAI proposed tougher supplier contracts to preserve operator control over AI risks
Operators would need stronger due diligence and disclosures for third-party AI embedded in telecom equipment or software. Supplier contracts could require regulatory compliance, , incident-management duties, risk allocation and accountability clauses, without transferring the operator's ultimate responsibility.
TRAI proposed human oversight, data governance and consumer safeguards for telecom AI
The draft policy framework would require human oversight of AI deployment and escalation where automated decisions could affect users or network services. It would also cover data governance, consumer protection and , including corrective action when systems create fairness or reliability concerns.
TRAI proposed central incident databases and mandatory reporting of critical and high-severity AI failures
Telecom entities would maintain , classify them by severity and follow procedures for investigation and mitigation. , including material malfunctions and recurring failures, would have to be reported to TRAI in a prescribed format and timeframe.
TRAI proposed validation and impact assessments before AI deployment or major changes
The draft would require validation, conformity and impact assessments before a new AI system is deployed or an existing system is materially modified. Assessments would cover accuracy, robustness, explainability, fairness and risks such as .
TRAI proposed high-, medium- and low-risk classifications for telecom AI systems
Operators would classify every AI system according to its importance to network operations, effects on financial and operational outcomes, and potential impact on consumers. Systems would also be recorded in a and cleared through the operator's designated governance process before use.
TRAI proposed board-approved AI policies and governance committees for telecom operators
Each covered entity would need an internal AI policy approved by its board or equivalent governing body. The policy would address deployment approvals, risk classification, incident management, human oversight, data governance and , while an would supervise implementation.
TRAI proposed making telecom entities accountable for every AI outcome
The draft would hold accountable for the outcomes of AI systems they use. That responsibility would apply whether systems are built internally, procured from third parties, or developed through a combination of both.
TRAI opened consultation on draft responsible AI guidelines for telecom networks and services
released . The proposal is open for stakeholder comments until November 27, 2026, and is not yet final law.