TRAI proposes board-approved AI policies mandatory reporting of serious incidents by telcos

     New Delhi, Oct 30 (PTI) TRAI on Friday proposed a comprehensive draft framework for the responsible use of artificial intelligence in telecom networks that would require service providers to adopt board-approved AI policies, establish dedicated governance and safety committees, and report serious AI-related incidents to the regulator.

     The draft guidelines, released for stakeholder consultations, also seek to make entities accountable for the outcomes of AI systems deployed in their networks and services, "irrespective of whether AI systems are developed internally, procured/ outsourced from a third party or a combination thereof."

     Telecom Regulatory Authority of India (TRAI), in a release, said it has floated the "draft 'Guidelines on Responsible Use of Artificial Intelligence (AI) in Telecommunication Networks and Provision of Telecommunication Services' for stakeholders' comments".

     A look at the consultation paper shows that under the proposed framework, telecom operators and other authorised entities using AI will have to classify their AI systems into high-, medium- and low-risk categories based on their significance to network operations, financial and operational outcomes, and potential implications for consumers.

     Each entity will be required to formulate an internal AI policy, approved by its board of directors or equivalent governing body, covering risk classification, deployment approvals, incident management, human oversight, data governance and consumer grievance redressal.

     "All AI systems shall be subject to an appropriate human oversight for risk mitigation as per a predefined mechanism," the consultation paper said.

     The regulator has proposed that AI-related incidents be classified into four categories, namely critical, high, moderate and low severity, with critical and high-severity incidents mandatorily reported to TRAI.

     "Critical severity AI incidents and high severity AI incidents shall be informed to TRAI in a prescribed format, manner and timeframe," it said.

     Telecom companies will also have to maintain central databases of AI incidents and consumer complaints involving AI-related harm, with mechanisms for investigation, mitigation and resolution.

     The sector regulator noted that, at present, AI is increasingly being integrated into telecommunication networks and provision of telecommunication services.

     "With a view to ensuring consistent and responsible implementation of the India AI Governance Guidelines in the telecommunication sector, and to increase clarity, predictability and accountability in the ecosystem, TRAI has decided to frame guidelines on responsible use of AI in telecommunication networks and provision of telecommunication services after following stakeholders' consultation," TRAI said.

     TRAI has invited stakeholder comments on the draft guidelines by November 27.

     Significantly, the draft mandates the establishment of an AI Governance and Safety Committee (AGSC), headed by a senior executive designated as the AI Governance and Safety Officer, to oversee AI deployment and identify vulnerabilities, fairness concerns and emerging systemic risks.

     The committee will review AI systems at least once every financial year and ensure corrective action, including modification or decommissioning of systems where necessary.

     "AGSC shall periodically review - at least once in a financial year, and additionally whenever such a review is necessary - the performance of the deployed AI systems, including AI incident records and other relevant information to identify significant vulnerabilities, oversight weaknesses, fairness concerns, and emerging or systemic risks.

     "Based on such reviews, AGSO shall ensure that appropriate corrective measures, including modification, decommissioning of relevant AI system(s), where necessary, are undertaken," the consultation paper said.

     The guidelines further propose mandatory validation, conformity and impact assessments before deploying new AI systems or modifying existing ones, covering accuracy, robustness, explainability, fairness, resilience against model drift and human oversight.

     Operators will also be required to maintain an inventory of all AI systems and undertake due diligence before procuring AI-enabled solutions from third parties.

     "The applicable authorised entity shall maintain a comprehensive, up-to-date inventory of all active, inactive and decommissioned AI systems. The record of each AI system in the inventory shall include the source, purpose and risk classification of the AI system," it said, emphasising that no AI system shall be used unless it is a part of the inventory.

     TRAI said increasing integration of AI into telecom networks could enhance efficiency but also introduce risks from unintended model behaviour, biases and adversarial attacks, potentially affecting critical infrastructure resilience.

     TRAI said the guidelines on responsible use of AI in telecom will be applicable to any entity (applicable authorised entity) that holds a licence under the Telegraph Act, 1885, or an authorisation under the Telecommunications Act, 2023, for providing telecom services or for establishing, operating, maintaining or expanding communications networks, and "using or intending to use AI systems in telecommunication networks or services, or both".

(This story has not been edited by THE WEEK and is auto-generated from PTI)