Industry Outlook: Telecoms & Connectivity — Week of September 7, 2026

Market Outlook

• AI networking surge outpaces optical supply. Ciena expects its backlog to exceed 10 billion dollars by year end as AI data center and interconnect demand outruns supply chain capacity. Strong revenue but a skeptical market signals that investors doubt vendors can translate AI demand into predictable, margin-safe delivery. Telcos planning 400G/800G and ROADM refreshes face longer lead times and higher execution risk for AI-adjacent transport builds.
• BEAD funds pivot from satellites back to fiber. NTIA guidance confirms that so‑called non‑deployment BEAD funds will be recycled into further broadband deployment rounds, while earlier program choices favoring low‑cost satellite left roughly 21 billion dollars unallocated. States like Louisiana are already moving, targeting 15,000 BEAD locations by early 2027 as local ISPs ramp builds. Fiber and fixed wireless providers now see a second wave of addressable subsidy, but scrutiny on sustainability and mapping accuracy is tightening.
• Philippine telcos commit 2.2B dollars for 2026 capex. Operators in the Philippines have earmarked 2.2 billion dollars for infrastructure next year, framed as a vote of confidence in national digital transformation. Spend is expected to tilt to 5G coverage, fiber backhaul, and enterprise connectivity, with regulators keen on competition and rural reach. Global vendors and towercos see another high‑growth Southeast Asian market leaning into 5G and fiber, but price pressure will stay intense.

Discussion: CTOs should assume continued tightness in optical and routing supply for AI‑heavy builds and factor BEAD’s second deployment wave into North American demand planning. Watch how subsidy design, especially around satellite versus fiber, reshapes regional competitive dynamics through 2027.

Headwinds

• AI agents expose flaws inside 4G and 5G cores. Recent research using AI agents uncovered dozens of previously unknown vulnerabilities in 4G and 5G core networks, many on internal interfaces that operators historically treated as trusted. The findings show that control plane and east‑west traffic inside the core can no longer be assumed safe, particularly as networks adopt cloud‑native microservices and third‑party components. Regulatory and enterprise customers will start asking pointed questions about internal segmentation and zero‑trust adoption in mobile cores.
• Supply constraints threaten AI transport timelines. Ciena’s growing backlog, despite strong revenue, highlights structural constraints in the optical and coherent component supply chain. AI data center interconnects are absorbing capacity that telcos expected to use for metro and backbone upgrades, and investors are already punishing vendors for execution risk. Operators that rely on single‑vendor optical stacks or just‑in‑time procurement face higher odds of deployment delays and rushed, suboptimal design choices.
• SmarTone revenue slide flags 5G monetization strain. Hong Kong’s SmarTone has warned it must substantially strengthen its business as revenue pressure mounts, despite extensive 5G rollout. Mature urban markets with near‑universal 4G and intense price competition are struggling to turn 5G capex into differentiated ARPU growth. That pattern signals risk for operators that treat 5G as a coverage race without parallel investment in enterprise services, slicing, and network APIs.

Discussion: Defensive moves should include accelerating zero‑trust and observability inside the core, diversifying optical vendors and form factors, and tightening investment cases for 5G expansion that lack clear monetization paths. Expect more board‑level questioning of 5G and AI network ROI assumptions.

Tailwinds

• 6G and AI alignment opens new service classes. Qualcomm and MasOrange/Ericsson are publicly tying 6G roadmaps to AI‑native devices and services, with recent 6 GHz trials suggesting up to fourfold capacity gains using advanced antenna arrays. Vendors are positioning 6G as both a higher‑capacity layer and a sensing and compute fabric for digital twins, industrial automation, and advanced positioning. Early clarity on spectrum candidates and AI‑centric use cases gives operators a narrative to engage regulators and large enterprises well before commercial 6G.
• HAPS and non‑terrestrial networks near commercialization. SoftBank has completed a stratospheric HAPS trial off Cape Muroto, validating connectivity between drones and edge computing systems over a high‑altitude platform and targeting deployment around 2027. Studies and industry panels continue to argue that terrestrial mobile networks will coexist with LEO and other non‑terrestrial systems rather than be displaced. That combination of HAPS, LEO, and terrestrial 5G opens room for multi‑layer coverage strategies in rural, maritime, and disaster‑prone regions.
• BEAD and state builds accelerate fiber footprint. New NTIA guidance that recycles unused BEAD funds into further deployment, plus visible progress in states like Louisiana, signals that the US broadband build cycle will extend into the late 2020s. Smaller operators and electric co‑ops are emerging as serious fiber players, often with open‑access or wholesale‑friendly models. Larger telcos can tap this expansion to sell backhaul, Ethernet, and NaaS to regional ISPs that lack core and edge capabilities.

Discussion: To capitalize, CTOs should start framing 6G and non‑terrestrial networks as extensions of current 5G and edge plans, not separate bets, and build product roadmaps for coverage‑as‑a‑service and NaaS offerings that ride on the BEAD‑driven fiber wave.

Tech Implications

• 6 GHz and 6G trials reshape spectrum strategy. MasOrange and Ericsson’s successful 6 GHz band tests, with advanced antennas and AI‑driven services, reinforce 6 GHz as a front‑runner for early 6G deployments in many regions. The trials suggest operators can achieve major capacity gains without immediately jumping to much higher, more fragile bands. Spectrum policy debates will intensify around licensed versus unlicensed 6 GHz, so engineering teams need clear positions on coexistence, interference management, and device ecosystem timelines.
• AI‑first networking drives optical and silicon design. Ciena’s backlog and Broadcom’s sharp networking revenue growth, tied to AI chips, show that AI workloads are now the primary design point for high‑end networking. Architectures are shifting to flatter, higher‑capacity fabrics, heavy east‑west traffic, and tighter coupling of optical transport with data center switching. Telco transport and edge designs that ignore AI data center interconnects and AI inference at the edge risk rapid obsolescence.
• Core security must move to zero‑trust by default. AI agents finding flaws in 4G and 5G cores highlight how traditional trust boundaries inside the core no longer hold, especially in multi‑vendor and cloudified environments. Network functions, APIs, and service meshes need mutual authentication, fine‑grained authorization, and continuous anomaly detection, not just perimeter firewalls and APN isolation. Vendors and operators that can show measurable hardening of internal interfaces will have an advantage in enterprise and government bids.

Discussion: Engineering teams should revisit spectrum roadmaps with a specific view on 6 GHz, redesign transport and peering to align with AI traffic patterns, and accelerate zero‑trust adoption in cores and edge clouds. Architecture choices made in the next 12 to 18 months will set the ceiling for 6G and AI service agility.

CTO Action Items

Prioritize a review of your optical and routing roadmap against AI data center and edge traffic forecasts, and assume component lead times will remain stretched through at least 2027. Direct your security and core engineering teams to treat all internal 4G and 5G interfaces as untrusted, with a concrete plan for mutual TLS, stronger identity, and deeper observability across control and user planes. Engage regulatory and spectrum teams on 6 GHz and non‑terrestrial integration, tying early 6G concepts to concrete AI and industrial use cases for your top enterprise accounts. Finally, map where BEAD or similar subsidy programs intersect your footprint, and identify where you should be a wholesaler, a NaaS provider, or an anchor enterprise partner rather than only a retail ISP.