As India’s 5G network matures and artificial intelligence (AI) moves from the experimentation stage to become an operational necessity, the two technologies are converging to redefine what telecom networks can deliver and who controls the intelligence layer of the digital economy. During a session on 5G and AI Convergence at tele.net’s flagship conference “5G & Beyond”, held in New Delhi, Debashish Chakraborty, Senior Director – Advocacy and Industry Engagement, GSMA; and Sandeep Sharma, Vice President and Head of Emerging Tech – Network Services, Tech Mahindra, discussed how this convergence is reshaping network operations, enterprise use cases and India’s broader AI ambitions. Edited excerpts…

Debashish Chakraborty

India’s 5G journey has crossed a significant threshold. As of April 2026, the country has over 410 million 5G subscriptions, a scale at which the technology begins to deliver its deeper potential. The conversation has moved beyond roll-out milestones. What matters now is what telecom networks are becoming – not connectivity pipes, but intelligent digital infrastructure that both powers and depends on AI.

Three developments are making this convergence structurally unavoidable rather than theoretically interesting. First, India’s 5G scale is now backed by growing device adoption, expanding fibre and data centre investments and accelerating enterprise digitalisation ambitions, all of which require low latency and high bandwidth. Second, AI adoption across Indian sectors is accelerating, reinforced by the IndiaAI Mission’s focus on compute ecosystems and data platforms. Third, and perhaps most consequentially, telecom networks have become too complex to manage manually. AI is now essential for predictive traffic management, fault detection, automated optimisation, energy efficiency and autonomous network operations.

This is not a future possibility but a present necessity.

The convergence is best understood through networks for AI and AI for networks. Networks for AI describes the role of connectivity as delivery and enablement infrastructure for AI workloads running on devices, at the edge, in the cloud or across distributed systems. Examples include edge infrastructure for low latency AI applications, 5G connectivity for AI-enabled internet of things (IoT) devices and AI-powered smart manufacturing and logistics systems. Meanwhile, AI for networks is where AI is deployed within the network to manage, optimise and operate infrastructure more intelligently than rule-based systems ever could. This covers autonomous network management, predictive traffic forecasting, dynamic spectrum allocation, self-optimising networks, predictive maintenance, automated fault detection and energy optimisation through smart sleep modes applied at the radio access network level.

The scale of what is already happening at the consumer layer illustrates the stakes clearly. AI usage alone is generating 77 exabytes of data per month globally, accounting for just 2-3 per cent of total wide area network traffic today. When that share reaches 30 per cent in the 6G era, networks will need to operate at an entirely different order of intelligence and capacity.

A GSMA survey of top global mobile network operators found that their three top AI priorities are improving customer experience, creating AI-driven new revenue streams, and improving network planning and operations. The next tier covers employee productivity and operational expenditure reduction. Sustainability and regulatory compliance follow. On compliance specifically, AI is enabling mobile operators to automate compliance monitoring, detect anomalies and violations in real time, strengthen fraud and scam detection, improve cybersecurity compliance, and monitor the quality of service and service level agreement performance continuously. These are not just marginal improvements. They address some of the costliest and reputationally damaging operational challenges the industry faces.

The convergence of AI and 5G also presents a significant energy challenge that the industry must confront honestly. Global telecom energy consumption stood at approximately 290 TWh in 2025, representing close to 1 per cent of total global energy use. By 2030, that figure is projected to reach approximately 500 TWh. A 5G base station already consumes anything between 1.5 to 3 times more power than its 4G equivalent. Although 5G base stations typically consume more power in absolute terms, 5G networks are considerably more energy-efficient per unit of data carried. By 2030, AI inferencing at the edge alone is expected to account for 35-40 per cent of total network energy demand.

The industry is not standing still. GSMA’s “Mobile Net Zero” report tracks climate commitments across the sector. The mobile industry has set a net zero target for 2050, while a number of leading global operators brought that target forward to 2040.

For GSMA, the work on AI is organised around three objectives.

  • Monetise: The billions invested in 5G networks must generate returns and AI is central to unlocking new revenue streams and demonstrating telcos as trusted providers of AI infrastructure.
  • Adopt: Scaling AI deployments requires industry collaboration, open-source support and the removal of adoption barriers.
  • Democratise: AI must work for all, including emerging markets and must be built on the principles of responsible use, sustainability and bridging the digital divide.
  • India’s own AI mission aligns closely with all three objectives and GSMA works directly with the programme as part of its broader commitment to the Indian market.

Sandeep Sharma

There is a framing for understanding why the convergence of 5G and AI is not simply a technology trend but a structural shift for the entire telecom industry. AI today is what English became for global business. A common language without which industries cannot communicate, cannot function and cannot compete. Every sector is grappling with this reality, but telecom carries a particular weight in that conversation. Of all the technologies shaping the modern world, telecom is the one that interfaces with almost every individual. A phone is more than a device. It is the user’s digital identity, reflecting what they think, what they plan and what concerns them. This makes telecom not only a conduit for AI but one of its most data-rich environments. The question is how to harness that data effectively.

The convergence of AI and 5G is currently delivering business value along two distinct dimensions.

  • The first is efficiency, specifically in decision-making. Telecom networks generate enormous volumes of data, but much of it has historically been fragmented, uncorrelated and difficult to process. AI brings the ability to process this data in real time and support faster, better-informed investment and operational decisions.
  • The second pillar is the creation of new use cases. Just as voice defined 2G and video streaming defined 4G, AI is the defining capability of the next generation. It is, in effect, the new voice that people need, something they will organise their digital lives around and something networks must be architected to support.

The honest challenge here is legacy. Networks have evolved across multiple generations and not all of them were built with AI-ready data exposure in mind. Older generations created data in formats that are difficult for AI systems to consume cleanly. As networks evolve, data accessibility will improve and AI can work with telecom infrastructure far more effectively.

Tech Mahindra’s work in this space centres on helping operators move from network automation to genuine network autonomy, and the distinction matters. Automation creates systems that work independently within defined parameters. Autonomy closes the loop. The system not only operates automatically but detects problems and takes corrective action without human intervention. The TM Forum has defined a maturity scale for this journey up to Level 5 autonomy. Most networks today operate at approximately Level 3, which means there is significant ground to cover. The path to Level 5 requires network transformation at multiple layers, including cloud-native functions, AI-ready data pipelines, intelligent orchestration platforms and redefined operational processes.

When it comes to which sectors will see the fastest adoption of AI-enabled 5G use cases in India, three categories stand out. The first is scale-driven sectors, where AI delivers its greatest value when operating at volume. Transportation, railways and supply chain management are strong candidates. The second is precision-and efficiency-driven sectors, particularly manufacturing, where autonomous processes and high accuracy requirements make AI integration compelling. The third is access to intelligence. Remote healthcare is a clear example, where AI can enable diagnostics and specialist-level decision-making in geographies where human expertise is scarce.

India is at a genuinely unique juncture. New digital infrastructure is being built from the ground up while the country’s workforce is rapidly developing AI capabilities. This combination of fresh infrastructure and AI-ready talent means that use cases developed and proven in India can be replicated across the world.

The reverse is not necessarily true. Solutions built for other geographies may not scale to India’s diversity and volume. With the IndiaAI Mission, the Quantum Mission, the Bharat 6G Alliance and the Semiconductor Mission all running simultaneously, the conditions for building an AI-native nation are in place.

Looking five years ahead, the most successful telecom operators will not be in the business of connectivity. They will be in the business of intelligence, using their networks as platforms to enable industries, generate data-driven revenue streams and deliver outcomes rather than simply provide connectivity.

The transition has already started. The operators that move decisively now will define what telecommunications will look like in the next decade.