In case there was any doubt, Nokia has jumped on the AI mystery train (an LLM loco-motive to cloud La La Land?), and also unhitched some of its difficult old rail cars in the process. It seems at once entirely logical, and completely mad – and follows on the heels of Nvidia’s $1 billion investment in the firm, which is (on paper) a vehicle to combine its twin AI strategies around fibre infrastructure (for hyperscalers et al) and wireless infrastructure (for telcos). So what have we got? At its Capital Markets Day in New York (November 19), Nokia said it will split its business into two operating segments to align with the “AI supercycle”.

These new units, active from January 2026, will deliver ‘network infrastructure’ and ‘mobile infrastructure’, respectively. The first will comprise optical networks, IP networks, and fixed networks, and is positioned as its “growth segment”; the second will combine its cellular core networks and radio networks portfolios (plus ‘technology standards’, formerly Nokia Technologies), and drive its 6G ramp-up – presumably as a post-2030 ‘growth segment’, while it gambles on AI infrastructure between times as telcos see out the last days of an underwhelming public-5G investment cycle. Seems fine, right – about in line with every other corporate strategy shift in recent times.?

What is mad – bonkers, even – is what it is dumping over the side: including its fixed-wireless access hardware (part of its existing ‘network infrastructure’ unit); its ‘enterprise campus edge’ solutions (its private 5G unit; part of its soon-to-be-defunct ‘cloud and network services’ division); and its microwave radio solutions (currently part of ‘mobile networks’). Clearly, RCR – or maybe just me – has a line here, having written about telco adventures in the Industry 4.0 game for seven or eight years. And RCR is following the same path editorially, of course; but there is grief and grievance at the bigger trend, and a whole bunch of head-shaking at Nokia’s move at the same time. 

All of these assets will be repurposed as ‘portfolio businesses’, pending a final decision on their “future direction” in 2026. And so they are in limbo. The Finnish vendor showed a slide at Capital Markets Day that re-calculated its 12-month 2024/25 run-rate if the new structure was already in place: the new divisions have delivered $7.8 billion (network infrastructure) and $11.6 billion (mobile infrastructure) in sales, respectively, on margins of 43% and 48% (and profit, somehow, of $0.8 billion and $1.5 billion). The rest have delivered a paltry $0.9 billion, at a loss of $0.1 billion. But wait; because Nokia is also about to quit a high-growth business, however niche, where it is king.

In private networks, Nokia is the heavyweight champ, right now – a double-digit upside since the start, in a market that will grow for 20-30 years, and which also blurs with a bunch of parallel value-generation opportunities (including AI hardware and software; useful edge placements for Nvidia GPUs). Whereas the big telecoms market (public networks) is totally flat – or pretty flat, anyway (1% growth, according to slides in New York). In private networks, Nokia has made the running, along with a few others – for the whole market. It has the lion’s share (upwards of 50 percent, reckon some analysts), and total authority. It has got to be mad to knock that on the head – if that is its plan.

Its broader enterprise business, cutting across various of its portfolio segments, is considerable, and hardly got a mention in New York – versus its twin cloud and telecoms strategy. Someone reflected: “Nokia is the only telco that has managed to diversify over the last five years – after the telco market started to decline. It has a strong enterprise business, overall, which cuts across most of its existing business segments, and a really strong-growth enterprise unit (in campus-edge / private-network sales) as well. It seems to be burying its best success stories.” Analysts that follow Nokia’s enterprise strategy are also bemused. 

Pablo Tomasi at Omdia told RCR Wireless: “It is accelerating to ride the AI train. I assume it has decided it has no patience for a long-term transformative market such as private networks. The challenge it faced in mobile networks probably did not help as this indirectly put pressure on its previous strategic bets – such as campus – to showcase short-term gains. To me, its new strategy is a retrenchment into its safe space of fixed and mobile connectivity. For the private networks market, it is a tough blow because Nokia has been instrumental in driving it.”

Leo Gergs at ABI Research said: “It is a consequence of wrong actions from the wider industry. It is frustrating to see. Nokia’s pivot to AI infrastructure, sidelining the enterprise edge, is symptomatic of a bigger telco failure: that this industry has fought turf wars – one technology versus another – instead of creating a unified connectivity layer that enterprises could trust… [And] ironically, AI was never part of the story, which backfired. Instead of using AI to make private networks simple and scalable, telcos waited for AI applications to magically appear. That was the mistake. This is not the end of enterprise cellular, but it is the last wake-up call.” 

We will pick up with them tomorrow – in a proper news piece. 

Meanwhile, here’s a separate message from the team to ask for your opinion about quantum computing: While it is hard to pin down exactly when quantum computing will threaten today’s encryption technologies, it is clear that it eventually will. With leading operators taking steps today, we want to know how you are preparing for Q-day. Join your peers in responding to this survey to help us gauge quantum readiness.