From 5G rollout to AI-native networks: India’s next shift

At the Voice & Data 5G+ Conference, a panel discussion on Digital Infrastructure for Viksit Bharat examined how India’s telecom networks must evolve from large-scale 5G rollouts to AI-enabled, intent-driven network futures. Titled From 5G to AI-enabled Network Futures: Automation, Predictive Analytics, Intent-based Operations, Resilience and AI-native Readiness, the session brought together senior leaders from telecom, IT services and network technology firms.

The discussion was moderated by Sunil David, a digital technology consultant and former regional director for IoT at AT&T in India and Asia. Setting the context, David noted that India’s 5G debate has largely moved past coverage. “The conversation is no longer about how much of India has 5G,” he said. “The real question is what outcomes 5G is delivering, sustainably and at scale.”

David argued that the future of networks would not be defined by successive generations such as 5G Advanced or 6G, but by how deeply artificial intelligence is embedded into network operations. AI, he said, must move from being an add-on analytics layer to becoming central to closed-loop, AI-native operating models.

AI-native networks and architectural choices

Arvind Bali, Vice Chairman and Board Advisor at Globe Teleservices, highlighted the shift from passive to intelligent networks. Drawing an analogy with surveillance systems, Bali said, “Speed and capacity alone don’t create value. Real productivity comes when networks become programmable, predictive and intelligent, not just reactive.”

He added that AI-enabled networks would increasingly prevent problems before they occur, rather than merely enabling post-event analysis, particularly in areas such as security, fraud detection and enterprise communications.

Panellists discussed how AI is already reshaping network operations in India, where operators must manage significant diversity across devices, geographies and usage patterns. AI-driven predictive and proactive operations are increasingly being used to forecast congestion, optimise performance and improve efficiency. At the same time, network traffic is changing, moving beyond video to include industrial telemetry, extended reality applications and edge workloads, requiring networks that can dynamically adapt.

Architecture emerged as a central theme in the discussion. The panel examined India’s mix of non-standalone and standalone 5G deployments, stressing that the real differentiator lies in whether networks are cloud-native, programmable and highly automated. Capabilities such as network slicing, orchestration and software-driven control were seen as essential for enabling enterprise use cases at scale. Security and resilience were also highlighted as critical, particularly in a country like India where networks must support emergency response and maintain continuity during natural disasters.

Sudakshana Laha, Head of Services for MSIT and ADM Cloud Software and Services at Ericsson, focused on what it would take to operationalise AI-native networks at scale. She argued that operators often remain overly focused on network efficiency rather than business outcomes. “Ultimately, revenue is realised only to the extent that OSS and BSS systems allow it,” she said. “AI must be embedded across IT, network and business layers if operators want to move from pilots to sustainable monetisation.”

Laha outlined a phased maturity approach, starting with automation and moving towards closed-loop autonomy and intent-based operations, driven by clearly defined business indicators.

From pilots to production: economics and deployment hurdles

A recurring concern was the slow transition of enterprise 5G deployments from pilots to full-scale production. Panellists pointed to several constraints, including spectrum policy limitations, operational complexity and the challenge of establishing clear returns on investment. While public 5G networks benefit from established operational frameworks, private 5G deployments often face difficulties around maintenance, on-site support and long-term cost justification.

Amit Sethi, Head of 5G Solutions at TCS, addressed why many enterprise 5G deployments remain stuck at the pilot stage. He pointed to spectrum policy constraints, operational challenges and unclear returns on investment. “Moving from proof-of-concept to minimum viable product is where most enterprises get stuck,” Sethi said. “Without clarity on long-term ROI, operational ownership and scale economics, mass deployment becomes difficult.”

He added that private 5G networks face practical challenges that public networks do not, including on-site maintenance, change management and enterprise readiness to invest beyond experimentation.

The role of alliances and outcome-based models

From an ecosystem perspective, speakers highlighted gaps in interoperability, platform readiness and collaboration between telecom operators, system integrators, cloud providers and technology vendors. While India has developed strong 5G infrastructure capabilities, panellists argued that delivering real outcomes will require deeper partnerships, stronger orchestration layers and a shift towards outcome-based business models rather than traditional metrics such as bandwidth or data consumption.

Amit Sinha Roy, Senior Advisor for Digital Ecosystems, Alliances and Innovation, emphasised the need for stronger ecosystem collaboration. While India has built robust 5G infrastructure, he argued that outcomes remain limited due to fragmentation across platforms and partners. “Enterprises are not looking to buy gigabytes or latency,” Roy said. “They are looking for outcomes, delivered through solutions that work end to end.”

Roy called for deeper collaboration between telecom operators, system integrators, hyperscalers and technology providers, alongside a shift towards outcome-based service-level agreements rather than traditional connectivity metrics.

Charting the path to AI-native operations

The session concluded with a consensus that India’s digital infrastructure journey is entering a new phase. With connectivity largely in place, the focus must now shift to intelligence, resilience and monetisation. Panellists agreed that AI-native networks, capable of sensing intent, predicting demand and acting autonomously, will be critical to supporting smart factories, digital public services and enterprise transformation as India moves towards its Viksit Bharat vision.

Over the course of the session, the panel addressed how Indian telecom operators, internet service providers and ecosystem partners can transition from rapid 5G rollout to AI-enabled, programmable and monetisable networks while keeping costs and energy efficiency in check. The discussion focused on four broad areas: the current state of India’s 5G journey and gaps in outcomes; what “AI-native” truly means across operations, architecture and organisational models; evolving network economics and monetisation strategies; and the role of trust and resilience.

Responding to a question on the biggest shift underway, panellists agreed that the transition is from reactive networks to intent-based, autonomous systems. AI-native networks, they said, must be able to sense application intent, predict demand, and act automatically, whether to guarantee ultra-low latency for telemedicine or manage massive IoT deployments in smart factories. Such intelligence, they argued, will form the backbone of India’s digital infrastructure as it scales towards 2047. 

As David summarised towards the close of the session, “The future network will not just carry data. It will understand why the data exists and how best to deliver outcomes from it.”