Beyond Connectivity: How AI and Edge Computing Are Reshaping Industrial Infrastructure
According to a recent Nasscom analysis, telecom operators are now betting on three pillars — AI automation, edge intelligence, and private 5G — to rewire how factories, utilities, and logistics firms handle data.

The enterprise networking industry is selling the same pitch it has been selling for a decade: that the next generation of infrastructure will solve the last generation's mistakes. According to a recent Nasscom analysis, telecom operators are now betting on three pillars — AI automation, edge intelligence, and private 5G — to rewire how factories, utilities, and logistics firms handle data. The question worth asking is who audits what gets wired up along the way.
The intelligent network is also a wider one
The Nasscom analysis frames telecom networks as evolving from "communication pipelines" into "intelligent digital platforms capable of processing, optimizing, securing, and orchestrating massive amounts of data in real time." Strip the marketing varnish, and this means more compute nodes, more orchestration layers, more autonomous decision-making — all sitting closer to physical operations than ever before.
Each new compute node is a new entry point. Every AI model that optimizes traffic routing is a model that can be poisoned. Every edge gateway that handles industrial telemetry is a gateway a threat actor can pivot through. The blast radius of a compromised orchestration plane now extends to assembly lines, substations, and ports rather than mail servers.
The report's own framing undercuts its optimism. It notes that AI workloads are "increasingly distributed" and that telecom operators are expanding their role precisely because there is data to be processed at the edge. Distributed data means distributed risk. Vendors will sell AI-native orchestration; defenders inherit a lateral movement problem with a much larger map.
Sovereignty, repackaged
The Nasscom piece cites a Financial Express op-ed arguing that "open networking ecosystems and indigenous telecom platforms can help nations build auditable, interoperable, and secure next-generation telecom infrastructure." The phrasing is careful. Auditable. Interoperable. Secure. Three adjectives, three different governance regimes, none of them enforced by the same body.
"Sovereign" infrastructure is being marketed as a response to "global restrictions around semiconductors, AI systems, and telecom technologies." Translation: when supply chains tighten, nations want their own stack. But owning the hardware and owning the threat model are different problems.
A sovereign 5G core running on domestic silicon does not stop credential reuse, unpatched baseband vulnerabilities, or supply-chain compromise baked into the radio units themselves. It moves the trust boundary. It does not eliminate it. And it tends to inherit the procurement opacity of any large government program.
The legacy floor
The most uncomfortable line in the Nasscom analysis is almost a throwaway: many utility operators "still depend on fragmented combinations of legacy infrastructure and public carrier networks." This is the floor under the entire Industry 4.0 pitch.
Private 5G is supposed to deliver deterministic performance, support thousands of sensors, and minimize congestion. In practice, it gets bolted onto grids and substations that were never designed for IP-based telemetry. The new private network becomes one more network to defend — while the old SCADA stack keeps running because nobody is willing to take it offline.
That is the real lateral movement problem. Not the sleek new 5G core. The decade-old serial-connected relay behind it. The infrastructure is getting smarter. The governance is not catching up. That gap is where the next incident will live.