Follow us on Google News
Telecom operators may be held accountable for third-party AI failures under proposed rules issued by India’s telecom regulator.
The Telecom Regulatory Authority of India (TRAI) has proposed a framework that would make telecom operators responsible for failures involving artificial intelligence (AI) systems used in their networks, even when those systems are developed by external vendors. The proposal raises questions about how responsibility should be divided between telecom companies and AI developers, particularly as the legal position on autonomous AI systems remains unsettled.
The proposed guidelines, issued on October 30, 2026, are open for public comments until November 17. They would apply to AI systems used in telecom operations, including customer-facing services and automated network functions. Under the framework, operators would remain accountable for AI-related failures while being required to establish safeguards, oversight mechanisms and contractual arrangements with technology suppliers.
https://the420.in/fcrf-certified-chief-ai-officer-second-cohort-november-2026/
What Responsibilities Would Telecom Operators Have?
TRAI’s proposed framework requires operators to assess AI systems before deployment, maintain human supervision and establish procedures for investigating failures. Companies would have to examine the credibility of AI suppliers, the reliability and limitations of their systems, and the suitability of data used for training.
Operators would also need enforceable contracts covering regulatory compliance, inspection rights, incident management, risk allocation and accountability. These arrangements would determine how responsibility is shared when an externally developed AI system causes harm.
The proposal also calls for testing before deployment, regular performance monitoring and mechanisms to address consumer complaints. Each operator would be required to establish an AI Governance and Safety Committee headed by a senior executive. The committee would review AI systems at least annually and could order modifications or shutdowns.
The requirements could also affect AI-based spam detection systems. TRAI previously directed telecom operators to deploy AI-based filters for calls and SMS in 2023. Under the proposed framework, operators could face responsibility when such systems make mistakes.
Why Has Anthropic Raised Liability Concerns?
Anthropic, the developer of Claude AI models, has identified uncertainty surrounding AI liability as a significant business risk. In its registration statement filed with the US Securities and Exchange Commission on October 26, the company acknowledged that contractual protections might not fully protect it against legal claims.
The company highlighted unresolved questions about responsibility for autonomous AI agents, which can perform tasks and make decisions without direct human intervention. Existing negligence and product liability laws have not been conclusively applied to such systems, creating uncertainty for developers and customers.
Anthropic also warned that customers and third parties could seek compensation for harm caused by AI agents, including situations where customers configured or supervised those systems. Errors could accumulate across multiple automated actions, increasing potential damage before problems are detected.
The company identified additional concerns involving insurance coverage and contractual commitments. It has accepted unlimited liability under certain agreements involving intellectual property claims, while warning that major AI incidents could exceed available insurance capacity.
Can AI Vendors Meet Reporting Requirements?
A major difficulty involves maintaining records needed to investigate AI failures. TRAI’s draft expects operators to examine incidents and report relevant failures, but technology suppliers may not always retain the information necessary for those investigations.
Anthropic has disclosed that a substantial portion of its traffic operates under Zero Data Retention arrangements. Under these terms, customer prompts and outputs are deleted, while limited automated safety information is retained.
Such arrangements may complicate investigations, regulatory examinations and incident reporting. Anthropic has acknowledged that restrictions on record retention could affect its ability to provide information within regulatory deadlines.
The company has also questioned whether human supervision can remain effective when autonomous AI systems operate at speeds and volumes beyond the capacity of reviewers to assess individual decisions.
How Does the Proposal Fit India’s AI Framework?
TRAI’s proposal follows India’s broader AI Governance Guidelines, which recommend assigning responsibility to organisations developing and using AI according to their roles and levels of risk. The telecom proposal places primary accountability on operators deploying the technology.
The Ministry of Electronics and Information Technology is separately working on wider AI regulation. IT Minister Ashwini Vaishnaw indicated in October that a consultation paper on AI regulation would be released within a month. The ministry has also indicated plans to strengthen requirements for reporting AI-related incidents.
However, TRAI’s draft introduces no additional penalties. Existing laws, regulations and telecom licence conditions would govern violations involving AI systems.
Several practical questions remain unresolved, including how operators should respond when vendors refuse contractual liability, how deleted records would affect investigations, and how regulators would enforce oversight requirements.
The proposal highlights a central regulatory challenge: telecom operators may be required to answer for AI failures even when the underlying technology, operational records and autonomous decisions remain partly outside their control.
