Private Network
Intelligence
Intelligence Journeys
Utilities Private Broadband Planning
AI Use Case Planning for Utilities

Policy

The European Commission’s Digital Networks Act (DNA) is a sweeping proposal to harmonize telecom rules, catalyze next‑generation investment, and turn 27 national markets into a functional single market for connectivity. The DNA is timed to underpin an AI‑driven economy that depends on fiber, 5G/6G, and low‑latency cloud‑edge fabrics spanning borders. Longer licence durations and more flexible sharing are intended to reduce renewal risk and unlock investment in 5G densification and 6G prep. Mandatory national plans to phase out copper between 2030 and 2035 will free OPEX and energy, but require careful migration of regulated wholesale products, vulnerable users, and critical services.
More than 60% of U.S. households are now serviceable by fiber, with double‑digit millions of new homes added in 2025 as operators, co‑ops, and municipalities pushed into suburban and rural areas. Median build benchmarks show underground construction around $18 per foot and aerial around $8 per foot, based on a broad sample of operator and contractor data across dozens of states. Most builders expect costs to rise again in 2026, and many anticipate longer schedules. Labor and materials remain the top pressure points, while permitting and make‑ready feature more prominently as networks stretch into harder‑to‑build pockets.
Enterprises are moving fast to private 5G to digitize operations, but the payoff only materializes if security scales with the new connectivity footprint. Private 5G brings deterministic wireless to factories, hospitals, ports, and energy sites, connecting robots, AGVs, cameras, and critical control systems. Security must follow identities and workloads, not subnets. Adopt a Zero‑Trust approach aligned to NIST SP 800‑207 with a single source of truth for identity and policy. Shift from perimeter controls to context-driven segmentation. Build on open standards and APIs to avoid lock‑in and simplify operations. Security must be foundational, measurable, and auditable from day one.
AI-driven experiences are flipping the traffic mix, pulling more capacity demand toward the uplink than U.S. mobile networks have historically planned for. Generative and vision-based AI are shifting usage from predominantly downloads to more continuous and bandwidth-heavy uploads. Recent benchmarking shows U.S. 5G networks prioritize downlink KPIs more than peers in Asia, even as uplink usage climbs. RootMetrics’ drive testing in late 2025 found all three U.S. carriers set roughly one-fifth of their midband Time Division Duplex (TDD) frame resources for uplink. That gap becomes material as AI, livestreaming, and enterprise camera workloads expand. U.S. carriers continued to win experience awards in early 2026, even as their uplink allocations trailed global leaders.
A new analysis of U.S. fixed wireless access shows subscriber momentum outpacing performance, a signal that capacity and management strategies are under pressure. Fixed wireless access from T-Mobile, Verizon, and AT&T added about 1.04 million net customers in Q3 2025, taking the U.S. FWA base to roughly 14.7 million. T-Mobile remains the U.S. FWA speed leader, posting a median download around 209 Mbps in Q3 2025. Median uploads dipped below 20 Mbps across providers, creating a hurdle for the FCC’s 100/20 benchmark. Urban FWA customers are more likely to meet the FCC’s 100/20 threshold than rural users due to radio geometry and site density.
Across roughly 2,000 decision-makers in telecom, data center, and large enterprises, a strong majority doubts that existing infrastructure will keep up with AI’s next wave. In the US, most respondents expect network buildouts to lag AI investment and call out near-term priorities such as optimizing bidirectional data flows, expanding fiber capacity, enabling real-time training feedback, and placing low-latency compute closer to users. In Europe, most enterprise leaders say current networks are not ready for broad AI adoption; many already report latency, throughput, and resiliency pain as data demands rise. The common thread is clear: without accelerated modernization, networks risk becoming the bottleneck that constrains AI outcomes.
A European 6G-XR consortium led by Capgemini, Ericsson, i2CAT and Vicomtech demonstrated holographic calling and edge-anchored XR services on live standalone 5G, signaling how networks will evolve to support immersive collaboration at 6G scale. The team executed end-to-end trials of real-time holographic communication and distributed XR experiences spanning edge nodes across Barcelona and Madrid. To keep spatial media stable under cell load, the partners implemented proactive congestion detection and an on-demand quality mechanism that prioritizes holographic traffic. Notably, the consortium has referenced IMS Data Channel as a vehicle to anchor real-time holographic streams within operator service frameworks.
UK regulator Ofcom has opened formal investigations into BT (including EE) and Three after nationwide voice service outages this summer impaired access to other networks and to emergency services. BT notified Ofcom of a software-related failure that disrupted interconnect voice services to and from the EE mobile network on 24–25 July 2025. During the incident, many BT and EE customers could not complete mobile calls to other networks or reach emergency services. Three separately reported a UK-wide disruption to voice services on 25 June 2025, which also affected some customers’ ability to contact emergency services.
As enterprises move from single-model chatbots to collaborative multi-agent systems, the economic and operational burden of reasoning at scale is becoming the dominant constraint. NVIDIA’s Nemotron 3 family introduces open models and tools designed to keep multi-agent systems fast, affordable and inspectable. The models use a hybrid latent mixture‑of‑experts design to activate only a fraction of parameters per token, combining it with a Mamba‑Transformer approach optimized for long sequences. Nemotron 3 Nano is a small, roughly 30B‑parameter model that activates up to 3B parameters per token, making it efficient for retrieval, summarization, assistants and software debugging.
Google has introduced a sharply priced AI Plus subscription in India to push generative AI into the mass market and counter OpenAI’s ChatGPT Go. The AI Plus plan launches at ₹199 per month for new users for six months, then moves to ₹399 per month. The bundle raises usage limits for Gemini 3 Pro, unlocks video generation within Google’s apps, expands NotebookLM’s “deep research” capabilities, and adds 200GB of storage across Google Photos, Drive, and Gmail. Family sharing is included, signaling a household-centric growth strategy.
Rogers Communications has moved from beta to a commercial footprint for satellite-to-mobile in Canada, extending basic connectivity and select apps to consumer smartphones while adding an industrial IoT tier for remote operations. The new Rogers Satellite service enables a curated set of popular apps to work beyond terrestrial coverage, including WhatsApp calling, Google Maps, AccuWeather, X, and CalTopo on most modern smartphones. In parallel, Rogers introduced satellite-to-mobile for IoT businesses, targeting asset tracking along highways and rail, as well as sensor telemetry in forestry, mining, and other resource sectors where terrestrial cellular is sparse.
The latest 3GPP cycle consolidates 5G-Advanced (Release 18), sets the agenda for Release 19, and frames early 6G studies, with direct implications for operator investment, enterprise use cases, and vendor roadmaps. Release 18, the first phase of 5G-Advanced, is moving from standards completion into implementation, bringing carrier-grade enhancements in spectral efficiency, energy savings, and mobility performance. Release 19 now builds on these foundations, prioritizing further RAN efficiency, XR and video delivery at scale, richer analytics/exposure in the core (NWDAF, NEF/CAPIF), and enterprise-grade reliability across private and public networks.

Partner Hubs

Telco AI Agent Leadership

Telco AI Agent Leadership Landscape - Network Equipment Vendors
Network Equipment Vendor Report
How are Amdocs, Salesforce, Comarch, Totogi, Qvantel, Oracle and other OSS/BSS vendors deploying telecom AI agents?
OSS/BSS Vendor Report
How are Amdocs, Salesforce, Comarch, Totogi, Qvantel, Oracle and other OSS/BSS vendors deploying telecom AI agents?
System Integrator Report
Scroll to Top