Thesis
Nokia is trying to become one of the less obvious winners of the AI bandwidth build-out. The key is that artificial intelligence does not only need more compute. It needs much more networking: inside the data center, between data centers, across metro and long-haul fiber, and increasingly across programmable telecom networks that will carry AI-driven services. Nokia sells into several of those layers at once through IP routing, optical transport, data center switching, network automation, and the APIs that make carrier networks consumable by software and enterprise agents. That is a very different investment case than the old "slow telecom equipment vendor" label still attached to the stock.
The setup is improving, but it still needs proof. In 2025, Nokia generated EUR 19.9 billion of net sales, EUR 2.0 billion of comparable operating profit, and EUR 1.5 billion of free cash flow. In Q4 2025, Network Infrastructure delivered EUR 2.407 billion of revenue, with 17% growth in Optical Networks, while management said order intake was strong across Optical and IP Networks with book-to-bill above one, driven by AI & Cloud customers. Nokia's 2025 annual report also said it generated EUR 2.4 billion of orders tied to AI and cloud customers. The real question now is whether that AI-and-bandwidth exposure becomes large enough to change how the whole company is valued, or whether the stronger network businesses remain trapped inside a broader Nokia story that still trades like a mature telecom supplier.
Nokia clearly has real exposure to rising AI-related bandwidth and network automation demand. The real question is whether those faster-growing businesses become important enough to overpower the market's old telecom-vendor framing.
Valuation and financials
The 4Ps
Justin Hotard is explicitly framing Nokia around connectivity for the AI era rather than around a generic telecom recovery. That matters because this is now as much a capital-allocation and portfolio-shaping story as a pure product story. Investors need management to keep steering resources toward AI & Cloud, optical, IP, and automation where the growth is clearly better.
Nokia's opportunity is not in one hero product. It sits across optical transport, IP routing, data center switching, automation, and network APIs. That breadth can be a strength because AI traffic growth does not show up in only one place. It hits back-end fabrics, front-end inference paths, inter-data-center transport, and the software layers used to orchestrate and monetize those networks.
The strongest bull case is that Nokia is no longer judged mainly on mobile-radio cycles. If Network Infrastructure, AI & Cloud, Infinera-enhanced optical, data center switching, and network-as-code software keep becoming more important, the mix can look structurally better and the multiple can improve with it.
Nokia is not as clean or concentrated as names like Ciena or Lumentum, and that is both good and bad. It gives the company more ways to win, but it also means the market needs more evidence before granting a premium. The confidence band is decent because multiple businesses are already profitable, but the rerating case still depends on mix improvement.
Portfolio manager lens
Starting point: Nokia is an underappreciated way to own the bandwidth and automation side of the AI build-out without paying for a pure-play data-center glamour multiple.
What is in the stock: stronger Network Infrastructure demand, optical and IP order strength, Infinera-enhanced AI transport exposure, and the idea that Nokia can become more relevant to AI & Cloud customers.
What can still surprise upside: AI & Cloud becoming a larger share of revenue and profit than investors expect, data center and DCI networking wins compounding, and network APIs plus autonomous-network software proving economically more meaningful than the market assumes.
What changes the view: AI-related wins remaining too small relative to the whole company, optical share gains not converting into better economics, or the old telecom-vendor framing continuing to dominate despite better strategic positioning.
Trade framing
This is not a clean pure-play AI-networking stock, which is exactly why it can still be interesting. The market already knows Nokia has exposure to networking. What it may still be underestimating is how much AI-driven bandwidth demand and network programmability can change the value of the better pieces of the portfolio.
The next checkpoints are practical: does Network Infrastructure keep showing strong Optical and IP order intake, do AI & Cloud wins and data center products keep scaling, and does management continue to simplify the story around fewer, clearer priorities? If yes, the stock can work as a gradual rerating rather than a single-quarter breakout. If not, it risks remaining a respectable but perennially discounted telecom name.
What matters now
What matters now is whether AI-driven bandwidth demand and network software and API monetization become big enough to outweigh the old telecom-vendor framing. The checkpoints are optical and IP-routing growth, data-center networking traction, and whether Network as Code and automation become real revenue lines.
Key questions
Nokia is broader than many investors remember. In 2025, it still reported through four business groups: Network Infrastructure, Mobile Networks, Cloud and Network Services, and Nokia Technologies. But management has already simplified the operating model beginning in 2026 around Network Infrastructure and Mobile Infrastructure, which is a clue to where the strategic emphasis is going.
For this thesis, the center of gravity is Network Infrastructure: optical transport, IP routing, fixed networking, and data center networking. That is the part of Nokia that benefits when AI workloads force more traffic across fiber, require lower-latency interconnect, and make data center and transport networks more programmable. Cloud and Network Services matters too because it includes the software, APIs, and automation layers that increasingly turn networks into programmable platforms rather than static plumbing.
Thesis last reviewed April 3, 2026. Live data updates automatically.