Telecoms Industry Assets Are Being Revalued in the AI Era and Not Everything Makes the Cut

Nokia has agreed to sell its Fixed Wireless Access (FWA) Customer Premises Equipment (CPE) business to Inseego in a transaction announced on April 30, 2026. The deal transfers a business generating approximately US$200 million in annual revenue in exchange for an estimated 11% equity stake in Inseego, alongside a small co-investment from Nokia at closing. Nokia and Inseego have also established a forward-looking collaboration framework around 6G and Artificial Intelligence (AI)-enabled wireless edge, meaning the relationship is converting from ownership to partnership rather than full separation. The acquisition roughly doubles Inseego’s revenue base and extends its footprint into Nokia's global operator ecosystem.

The divestment is the first major exit from Nokia's "Portfolio Businesses" segment, a category created at its November 2025 Capital Markets Day by Chief Executive Officer (CEO) Justin Hotard, who joined from Intel's Data Center and AI Group in April 2025. That segment was explicitly designed to house assets identified as non-core through a strategic review—units generating close to €0.9 billion in combined revenue but operating at a net loss. Nokia's restructuring has concentrated resources on two operating segments—Network Infrastructure and Mobile Infrastructure—targeting €2.7 to €3.2 billion in comparable operating profit by 2028, anchored in optical, Internet Protocol (IP) networking, mobile core, and data center interconnect.

Across the telecoms industry vendor landscape, a reassessment is underway of where value will ultimately lie in the full AI stack. Nokia's move is an early, visible inflection point in what is becoming a sector-wide evaluation of where vendor capital belongs in an AI-defined infrastructure cycle.

Nokia's decision to exit FWA should not be read as a verdict on the technology. FWA remains a commercially relevant and growing access category, particularly in markets underserved by fiber and in enterprise campus and backup connectivity scenarios. The global FWA subscriber base continues to expand, and spectrum availability in mid-band and Millimeter Wave (mmWave) ranges is improving the economics of deployment. The technology is not in decline. What has changed is how major telcos assign strategic value to the assets they hold, as well as attempting to reposition their assets for AI. Nokia's move is best understood as part of a broader shift in how telecoms industry vendors manage portfolios under AI-driven capital allocation pressures. Rather than maintaining wide product coverage as a competitive moat, vendors are increasingly concentrating investment in fewer domains with higher long-term margin potential and direct alignment with infrastructure-level value creation in the AI era.

The dividing line being drawn across vendor portfolios is proximity to AI-driven infrastructure value. Nokia is concentrating on network layers that interact directly with data center traffic flows, transport efficiency at scale, and AI workload demands—optical systems, IP routing, and mobile core. These are the layers closest to where the economics of AI infrastructure are being decided. CPE-heavy businesses like FWA, which sit closer to the connectivity delivery layer, face a structural ceiling on differentiation and margin expansion that vendors with AI-era ambitions are increasingly unwilling to manage. Purpose-built hardware sold into a market where operators hold significant pricing power, and where software-defined alternatives are advancing, does not compound value in the same way that orchestration and transport-layer infrastructure does.

Nokia is unlikely to be the only vendor drawing these boundaries. The same AI capital allocation pressures that drove this divestment are present across every major Tier One telco equipment vendor, and the internal portfolio logic Nokia has made explicit—concentrate on AI-adjacent infrastructure, externalize what sits outside it—is not unique to Nokia's strategy. Vendors across the sector are under the same investor pressure to demonstrate margin trajectory over revenue breadth. Nokia's FWA exit is a leading indicator of a broader rationalization cycle, not an isolated decision.

The appointment of Emma Falck as President of Mobile Infrastructure further reinforces Nokia’s shift away from a purely engineering-led, radio-centric strategy toward a more end-market and application-oriented view of mobile infrastructure. While this aligns with a broader repositioning toward AI-driven and enterprise-facing opportunities, it also introduces execution risk at a pivotal moment for 6G, where deep Radio Access Network (RAN) expertise and sustained Research and Development (R&D) investment remain critical to competitiveness and patent leadership. The move may signal that Nokia is recalibrating its expectations for 6G as a revenue driver relative to prior generations, prioritizing commercial relevance over technology leadership. Whether this transition enables differentiation or weakens Nokia’s position in a RAN-intensive 6G cycle will be a key fault line to watch.

Nokia sold its FWA business because it generated revenue in the wrong part of the stack. It was hardware-intensive, margin-constrained, and disconnected from the AI workload infrastructure that defines Nokia's forward-looking investment priorities. That characterization does not end with FWA. Enterprise campus networking and Internet of Things (IoT) gateway and device management businesses carry the same profile across multiple vendors, with commoditization pressure compressing margins and no natural convergence with optical, IP, or mobile core priorities. Stakeholders should expect additional exits from Nokia and others driven by the same underlying rationale.

Nokia’s exit reflects a clear prioritization of near-term, AI-adjacent infrastructure value. It also implies a reprioritization of more speculative, longer-term opportunities at the network edge. FWA and other in-home gateway technologies could evolve into platforms for consumer-facing AI services, where telcos embed intelligence directly into the home environment and position themselves as AI providers. However, realizing this vision would require sustained capital investment, ecosystem development, and significant go-to-market effort in what remains a nascent and uncertain market. Nokia’s decision suggests a deliberate trade-off: focusing on immediate, infrastructure-driven revenue pools, rather than pursuing longer-horizon, consumer AI platform plays.

More broadly, Nokia will not be alone. Ericsson, among others, has already been rationalizing its portfolio, having exited managed services and IoT segments over the past 2 years. What is emerging is not individual vendor housekeeping, but a synchronized rebalancing across the Tier One vendor landscape, driven by the same AI capital allocation logic and reinforced by investor pressure on vendors to demonstrate margin trajectory, not revenue breadth. The next 18 to 24 months are likely to produce several additional divestitures of this type across Nokia, Ericsson, and potentially others, as deliberate portfolio moves.

The ownership structure Nokia used here is also worth watching as a template. Rather than a clean cash sale, Nokia retained equity and a partnership role. This model allows vendors to shed capital intensity, while preserving upside in growing markets. It satisfies internal capital reallocation goals, while keeping the vendor embedded in adjacent ecosystems that may eventually converge with its core AI infrastructure play. Buyers should expect more minority-stake, partnership-anchored deals of this kind.

For operators and enterprises, the practical implication is that the telco supplier ecosystem is becoming more fragmented and more specialized simultaneously. The Tier One vendors that remain will be narrower in scope, but deeper in the specific infrastructure layers that matter most for AI-driven network operations. The specialists absorbing the divested categories will be larger and better resourced than their predecessors, but they will be optimizing for execution scale, rather than platform integration. Operators that currently rely on a small number of broad vendor relationships to manage their supply chain will face increasing pressure to build genuine orchestration capability across a wider and more fluid set of suppliers. That capability is a condition for maintaining operational resilience as this restructuring accelerates.

The underlying shift is structural and directional. Vendor portfolios are being actively engineered toward AI-relevant infrastructure, and the assets that do not qualify are being externalized. The question for every buyer is no longer whether their vendor will focus on AI—every major vendor will say that they are. The question is which specific business units sit inside their strategic view.