Indian telecom operators face liability for AI failures under proposed TRAI rules
The Telecom Regulatory Authority of India has proposed rules requiring telecom operators to assume full liability for the performance of all AI systems, including those sourced from third-party vendors.
The Telecom Regulatory Authority of India (TRAI) has introduced a draft policy mandating that telecom operators remain accountable for the outcomes of all artificial intelligence systems deployed within their networks. The proposed guidelines dictate that this responsibility applies regardless of whether the AI tools were developed internally or procured from external vendors.
Under the draft, operators must implement a board-approved AI governance policy, categorize systems by risk level, and maintain an inventory of all active AI tools. Furthermore, the regulation requires human oversight for every system and the creation of an AI Governance and Safety Committee. These bodies will be tasked with regular reviews and ensuring that incident reporting mechanisms are in place for the most severe system failures.
The industry faces potential legal friction as AI developers navigate their own liability. Anthropic, in recent SEC filings, noted that the legal frameworks governing autonomous AI agents remain unsettled. The company warned that contractual protections, such as warranty disclaimers, may be insufficient to limit their exposure to legal claims, potentially leaving telecom operators in a position where they cannot fully transfer risk to vendors.
TRAI is accepting public comments on the draft until November 27. If adopted, the rules will apply to all licensed telecommunications providers, shifting the burden of AI governance onto the network operators.