The Second Great Telecom Reckoning: Endurance in the Age of AI Dominance

By | January 30, 2026

It’s been a year and a half since I published my July 2024 paper (Mobile Infrastructure Capex: Permanent Weakening or Short-Term Decline) explaining why the telecom capex downturn would run deeper and longer than previous cycles. This piece aims to capture a current market snapshot by examining the most notable bankruptcies and market exits since that earlier publication. I believe the depth of this correction is second to the dotcom collapse. (Consolidation topics remain reserved for a future post.)

Market Exits

Echostar marked a major market exit for what was supposed to be the fourth U.S. network operator. The company sold its spectrum assets to AT&T and SpaceX in August and September 2025, raising more than $40 billion and covering its outstanding debt. US Cellular also exited, but because that announcement came before July 2024, I’m not including it here.

Mavenir aimed to become a major RAN vendor through OpenRAN. It exited the radio business in mid‑2025 as part of a $1.3 billion debt and equity restructuring plan.

Kyocera announced its exit from the base station and RAN market at the end of 2025, citing high development costs in Japan and weak profit prospects. The company had chosen an Nvidia‑GPU architecture, and its R&D budget sat at less than one‑tenth of what incumbents like Ericsson and Nokia spent.

NEC classified the RAN market as non‑core and exited by the end of 2025. Omdia ranked NEC as the sixth‑largest RAN vendor in 2024 with a 0.9% market share. NEC now plans to focus more on the defense sector, a direction that many telecom companies increasingly follow.

Microsoft pulled back from the 5G networking business in mid‑2024. It had acquired Affirmed Networks and Metaswitch in 2020 with plans to sell telecom functions to operators, but the strategy didn’t gain traction. Microsoft also scaled down Azure for Operators, including layoffs in June 2024. Reports in 2025 indicated that Alianza acquired Metaswitch assets from Microsoft.

AWS exited the private 5G market in mid‑2025. Its enterprise service included small‑cell radios, a 5G core, RAN software, and SIM cards. The offer was too expensive, and adoption remained slow. AWS cited limited spectrum availability and its reluctance to support third‑party hardware.

NXP exited the 5G RF power market and plans to wind down its power‑amplifier line by 2027. The company will close its GaN production facility in Arizona as part of the process (ECHO fab).

Commscope sold its connectivity and cable‑solutions business, which is its highest‑earning unit, to Amphenol to reduce debt. Amphenol paid $10.5 billion and expects the unit to generate $4.1 billion in 2026. Amphenol also acquired the Commscope brand. Commscope then rebranded as Vistance Networks, which now includes Access Network Solutions (renamed Aurora Networks) and Ruckus Networks, both of which Commscope had originally planned to divest. Commscope had already sold its Home Networks unit in 2023 as part of its ongoing decline.

Bankruptcies

Casa Systems filed for Chapter 11 in April 2024. Lumine Group acquired its 5G core and RAN business, while Commscope acquired the cable business.

Mitel filed for Chapter 11 in March 2025 to restructure $1.3 billion in debt. It emerged in June 2025 with roughly $160 million in debt.

Airspan filed for Chapter 11 in March 2024 after high R&D costs and 5G operating losses. The company converted funded debt into equity, eliminated existing debt, and raised $95 million in new financing. It emerged from bankruptcy in October 2024.

Satellite service provider Ligado filed for bankruptcy in January 2025 to reduce its $7.8 billion debt load to $1.2 billion. Ligado, formerly Lightspeed, had previously filed in 2012. It emerged again in December 2025.

Mobileum, a telecom analytics provider, filed for bankruptcy in July 2024 to restructure more than $500 million in debt. It reemerged in September 2024 with $60 million in new financing.

Concluding Thoughts

  • Sustaining four competitive mobile operators in a mature market is proving extraordinarily difficult, especially when the fourth entrant arrives late to an already saturated arena with high barriers and limited differentiation opportunities.
  • RAN vendors, particularly those betting on Open RAN to challenge the incumbents as well as Japanese vendors, suffered the most severe setbacks. This segment has borne the brunt of widespread capex pullbacks, exposing the fragility of diversification efforts in a contracting equipment market.
  • After years of aggressive pushes into telecom, cloud giants are decisively scaling back their ambitions. In the tug-of-war between traditional telecom infrastructure and the explosive demands of AI, AI has clearly emerged victorious, redirecting hyperscaler priorities and resources away from carrier-grade networking.
  • Semiconductor and infrastructure players are consolidating or exiting outright, a predictable outcome given the relentless pressures on equipment vendors: shrinking margins, stalled 5G ROI, and the broader industry shift toward efficiency over expansion.

The broader industry is in full consolidation/grind mode. RAN sales are still depressed (down sharply since 2022 peaks). Operators focus on cost-cutting, margin tweaks, and capital recycling (e.g., fiber/tower deals, spectrum divestments) rather than bold expansion. 2026 forecasts point to a modest revenue/EBITDA growth at best, driven by efficiency, not demand surges.