Industry Outlook: Telecoms & Connectivity — Week of August 24, 2026

Market Outlook

• Satellite D2D edges toward commercial reality. AST SpaceMobile secured temporary FCC approval to test direct-to-device services in 800 MHz with AT&T calling it a milestone for closing coverage gaps. Direct satellite-to-handset is shifting from experiment to roadmap item, which pressures mobile operators to define how satellite fits into national coverage, roaming, and emergency service strategies.
• Cable consolidation reshapes fixed access scale. Charter and Cox have wrapped their merger, alongside Liberty Broadband, creating the largest cable operator with about 70 million locations passed and 37 million customers under the Spectrum brand. That scale will matter for fiber upgrades, DOCSIS 4.0 timing, and wholesale backhaul, and it raises the bar for mid-tier ISPs that lack similar density and capital.
• AI traffic drives new capacity and politics. Operators and vendors report AI networks facing a triple constraint of capacity, permitting, and supply chains as GPU clusters proliferate across data centers. At the same time, AI data centers are becoming a political flashpoint in the US, which will shape where and how quickly hyperscalers and carriers can add power and fiber.

Discussion: CTOs should treat satellite D2D and AI-driven transport as near-term planning inputs, not long-term curiosities. Fixed-mobile convergence strategies need a fresh look in light of cable consolidation and rising wholesale dependence.

Headwinds

• AI-era networks hit capacity and permit walls. Lumen, AT&T and others highlight that AI workloads are pushing optical and metro networks toward capacity, yet new builds are slowed by local permitting and constrained supply chains. Without early re-architecture and vendor diversification, operators risk being the bottleneck for hyperscaler AI growth, weakening their negotiating position on long-term connectivity and colocation deals.
• Cybersecurity and copper retirements raise resilience risk. France’s SFR reportedly suffered a cyberattack, another reminder that large incumbents remain prime targets as they digitize operations. In parallel, Lumen and others are exiting wholesale telephony and retiring copper in marginal areas, which cuts legacy opex but can expose gaps in voice, regulatory obligations, and enterprise SLAs if migration paths are not engineered carefully.
• European mobile margins stay under pressure. Analysis of European telecom shows low-cost challengers like Iliad and Digi thriving while incumbents struggle to grow ARPU in markets addicted to handset subsidies. Vendor revenues are also under strain, which can slow R&D and support quality for key 5G and Open RAN programs and complicate long-term technology bets for operators.

Discussion: CTOs should stress test transport and IP core plans against aggressive AI traffic scenarios, and run structured risk reviews on cyber posture and copper or TDM retirement. Vendor consolidation and financial stress deserve explicit contingency planning in multi-year RAN and core roadmaps.

Tailwinds

• 5G SA and slicing show tangible performance gains. Verizon and Ericsson reported a 2.2x uplink spectral efficiency improvement on 5G standalone compared with non-standalone, a concrete data point for upgrading the core. EE launched its Fast Lane consumer and SMB slicing offer, joining a growing European cohort that is starting to turn 5G SA capabilities into differentiated products rather than generic capacity.
• Hollow-core fiber wins hyperscaler-scale order. Relativity Networks raised 22 million dollars and secured a 40 million dollar hyperscaler contract for hollow-core fiber, which promises lower latency and improved performance for AI and trading workloads. Hyperscaler adoption will pull advanced optical technologies into the mainstream and gives operators a reference for next-generation metro and data center interconnect design.
• AI and cloud segments start to pay off. Nokia reported a 105 percent year-on-year increase in AI and cloud segment sales to 446 million euros, now 9 percent of total revenue. Growing spend on AI-native network operations, analytics, and cloud-native infrastructure signals that operators are willing to fund software-driven efficiency and automation if vendors can show clear opex and capex benefits.

Discussion: CTOs can use the new 5G SA performance data and slicing offers as internal proof points to move stalled core upgrades. AI operations tooling and advanced fiber options should be evaluated now for inclusion in 2027–2028 capex plans tied to AI and low-latency services.

Tech Implications

• 6G cores likely to evolve from 5G, not reboot. Omdia analysts argue that while 3GPP debates an AI-native 6G core, suppliers cannot deliver a greenfield architecture by 2030 and operators cannot afford it. The first 6G cores will probably extend 5G SA designs with deeper AI and automation, which means investments in 5G cloud-native core, observability, and CI/CD will carry forward rather than be stranded.
• Agentic AI and open telco models need guardrails. NGMN sees agentic AI as a path to Level 4 autonomous networks, but flags identity, governance, interoperability, security, and cost control as prerequisites. GSMA adds that today’s frontier models do not understand telecom well enough, and calls for open telco-specific models, which shifts the AI conversation from generic LLMs to domain-tuned, standards-aware systems embedded in OSS and NMS.
• Edge computing repositions around AI inferencing. Ericsson argues the first wave of telco edge failed because workloads did not justify the footprint, but AI inferencing, robotics, and new network capabilities could change the equation. Telco edge nodes that are tightly integrated with 5G SA, slicing, and local breakout are now being pitched as inferencing sites for latency-sensitive enterprise and industrial applications.

Discussion: CTOs should align 5G core modernization, AI automation, and edge strategies as one program rather than three silos. Architecture choices now, especially around cloud-native tooling and data models, will decide how easily 6G features and autonomous operations can be adopted in the next decade.

CTO Action Items

Treat satellite direct-to-device as a concrete integration topic: assign teams to model coverage, roaming, and emergency service implications and to explore partnerships or MVNO-style agreements with emerging LEO D2D providers. Revisit your 5G SA and slicing roadmap using the latest Verizon and EE data as internal ammunition, and define at least one commercial slicing or differentiated QoS offer for 2027. Launch a focused review of AI-era transport and edge needs, including hollow-core fiber pilots for key metro or data center interconnect routes and a refreshed plan for GPU cluster connectivity. In parallel, tighten your automation strategy by selecting a small set of telco-specific AI tools, setting governance and security requirements up front, and ensuring your 5G core, OSS, and edge platforms expose the APIs that future network monetization and autonomy will depend on.