As AI‑native applications scale to more users, agents and devices, the telecommunications network is becoming the next frontier for distributing AI. 

At NVIDIA GTC 2026, leading operators in the U.S. and Asia showed that this shift is underway, announcing AI grids — geographically distributed and interconnected AI infrastructure — using their network footprint to power and monetize new AI services across the distributed edge.  

Different operators are taking different paths. Many are starting by lighting up existing wired edge sites as AI grids they can monetize today. Others harness AI-RAN — a technology that enables the full integration of AI into the radio access network — as a workload and edge inference platform on the same grid.  

Telcos and distributed cloud providers run some of the most expansive infrastructure in the world: about 100,000 distributed network data centers worldwide, spanning regional hubs, mobile switching offices and central offices, with enough spare power to offer more than 100 gigawatts of new AI capacity over time.

AI grids turn this existing real-estate, power and connectivity into a geographically distributed computing platform that runs AI inference closer to users, devices and data, where response and cost per token align best. This is more than an infrastructure upgrade — it’s a structural change in how AI is delivered, putting telecom networks at the center of scaling AI rather than just carrying its traffic. 

Global Operators Turn Distributed Networks Into AI Grids

Across six major operators, AI grids are moving from concept to reality.

AT&T, a leader in connected IoT with over 100 million connections across thousands of device types, is partnering with Cisco and NVIDIA to build an AI grid for IoT. By running AI on a dedicated IoT core and moving AI inference closer to where data is created, AT&T can support mission‑critical, real‑time applications like public‑safety use cases with Linker Vision, enabling faster detection, alerting and response while helping keep sensitive information under customer control at the network edge.

“Scaling AI services that are both highly secure and accessible for enterprises and developers is a core pillar of our IoT connectivity strategy,” said Shawn Hakl, senior vice president of product at AT&T Business. “By combining AT&T’s business‑grade connectivity, localized AI compute and zero‑trust security while working with members of the NVIDIA Inception program and harnessing Cisco’s AI Grid with NVIDIA infrastructure and Cisco Mobility Services Platform, we’re bringing real‑time AI inference closer to where data is generated — accelerating digital transformation and unlocking new business opportunities.”

Comcast is developing one of the nation’s largest low‑latency broadband footprints into an AI grid for real‑time, hyper‑personalized experiences. Working with NVIDIA, Decart, Personal AI and HPE, Comcast has validated that its AI grid keeps conversational agents, interactive media and NVIDIA GeForce NOW cloud gaming responsive and economical even during demand spikes, with significantly higher throughput and lower cost per token.

Spectrum has the network infrastructure to support an AI grid that spans more than 1,000 edge data centers and hundreds of megawatts of capacity less than 10 milliseconds away from 500 million devices. The initial deployment focuses on rendering high-resolution graphics for media production using remote GPUs embedded across Spectrum’s fiber-powered, low-latency network.

Akamai is building a globally distributed AI grid, expanding Akamai Inference Cloud across more than 4,400 edge locations with thousands of NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs. Akamai’s AI grid orchestration platform matches each request to the right tier of compute, improving the token economics of inference while powering low-latency, real-time AI experiences for applications like gaming, media, financial services and retail.

Indosat Ooredoo Hutchison is connecting its sovereign AI factory with distributed edge and AI‑RAN sites across Indonesia to build an AI grid for local innovation. By running Sahabat-AI — a Bahasa Indonesia-based platform — on this grid within Indonesia’s borders, Indosat can bring localized AI services closer to hundred millions of Indonesians across thousands of islands, giving local developers and startups a sovereign platform to build AI applications that are fast, culturally relevant and compliant by design.

T‑Mobile  is working with NVIDIA to explore edge AI applications using NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs, demonstrating how distributed network locations could support emerging AI-RAN and edge inference use cases. Developers including LinkerVision, Levatas, Vaidio, Archetype AI and Serve Robotics are already piloting smart‑city, industrial and retail applications on the grid, connecting cameras, delivery robots and city‑scale agents to real-time intelligence on the network edge. This demonstrates how cell sites and mobile switching offices can support distributed edge AI workloads while continuing to deliver advanced 5G connectivity.

New AI‑Native Services Put Telecom AI Grids to Work

AI grids are becoming foundational to a new class of AI‑native applications — real‑time, hyper‑personalized, concurrent and token-intensive.

Personal AI is using NVIDIA Riva to power human‑grade conversational agents on the AI grid. By running small language models closer to users, it achieves sub-500 millisecond end-to-end latency and over 50% lower cost-per-token, enabling voice experiences that feel natural while remaining economically viable at scale.

Linker Vision is transforming city operations by running real‑time vision AI on the AI grid. By processing thousands of camera feeds across distributed edge sites, it delivers predictable latency for live detection and instant alerting — enabling safer, smarter cities with up to 10x faster traffic accident detection, 15x faster disaster response and sub‑minute alerts for unsafe crowd behavior. 

Decart is redefining hyper‑personalized distributed media by bringing real‑time video generation to AI grids. By running its Lucy models at the network edge, it achieves sub‑12-millisecond network latency, enabling interactive video streams and overlays that adapt instantly to each viewer, delivering smooth, immersive live video experiences even when viewership peaks.

AI Grid Reference Design and Ecosystem

The NVIDIA AI Grid Reference Design defines the building blocks — including NVIDIA accelerated computing, networking and software platforms — for deploying and orchestrating AI across distributed sites.

A growing ecosystem of full‑stack partners including Cisco and infrastructure partners like HPE are bringing AI grid solutions to market on systems built with the NVIDIA RTX PRO 6000 Blackwell Server Edition. Armada, Rafay and Spectro Cloud are among the partners building an AI grid control plane to seamlessly orchestrate workloads across distributed AI infrastructure.

“Physical AI is accelerating the shift from centralized intelligence to distributed decision making at the network edge,” said Masum Mir, senior vice president and general manager provider mobility at Cisco. “Our partnership with NVIDIA brings together the full stack — from NVIDIA GPUs to Cisco’s networking and mobility capabilities — enabling operators to power mission-critical applications, deliver real-time inferencing and participate in the AI value chain.”

Together, this ecosystem is helping telcos and distributed cloud providers redefine their role in the AI value chain — transforming the network edge into a unified intelligence layer that runs, scales and monetizes AI workloads.

Learn more about AI Grid.