When senior leaders at Nokia (NOK -1.57%) start writing six-figure checks for their own stock, I pay attention. Over the past few months, a handful of Nokia executives and board members have quietly accumulated tens of thousands of shares, even as the stock has already enjoyed a strong run on the back of its artificial intelligence ambitions. Given who sits on the other side of those AI plans -- Nvidia (NVDA -0.10%) with a billion-dollar strategic investment -- this feels more like a deliberate bet than a casual perk.
NYSE: NOK
Key Data Points
Nokia insiders are loading up
In late May, Nokia disclosed that Victoria Hanrahan, chief of staff to the CEO, bought 44,682 Nokia shares in two New York Stock Exchange transactions at an average price of about $15.81 per share, a purchase worth just over $700,000. Then, on July 24, the company filed a managers' transaction report showing three more insiders buying: senior manager Patrik Hammarén acquired 43,293 shares in Helsinki at around 8.44 euros, board member Timo Ihamuotila picked up 60,000 shares across multiple European venues at roughly 8.45 euros, and senior manager Pallavi Mahajan bought 62,000 shares on the NYSE at about $9.55. These are not token purchases. They are meaningful personal commitments at prices that reflect the new, AI-focused Nokia rather than a turnaround bargain.
Image source: Getty Images.
The backdrop for that buying spree is Nokia's decision to tie its future networks directly to Nvidia's AI hardware. In October 2025, Nokia and Nvidia announced a strategic partnership to pioneer an AI platform for 6G, with Nvidia committing a $1 billion equity investment at a subscription price of $6.01 per share. The collaboration does two important things. First, it adds Nvidia-powered, commercial-grade AI RAN products to Nokia's existing radio access network portfolio, giving carriers a way to launch AI native 5G Advanced and 6G networks on Nvidia's new Aerial RAN Computer platforms. Second, it expands the partnership into data center switching and AI networking, combining Nokia's SR Linux software with Nvidia's Spectrum X Ethernet platform to optimize traffic inside AI clusters.
NASDAQ: NVDA
Key Data Points
Nokia is integrating AI into its process
Nokia is not just licensing a logo here. It is rearchitecting its base stations around Nvidia silicon. In a detailed announcement, Nokia laid out plans for AI RAN base stations that run all RAN processing on Nvidia GPUs, with no separate accelerator, and for Cloud RAN solutions that use the Grace CPU Superchip for higher-layer processing, while Nokia's in-line Layer 1 accelerator handles the physical layer. T-Mobile U.S. (TMUS +3.75%) has already agreed to trial these AI RAN designs in its networks, starting in 2026, which gives Nokia a real-world proving ground rather than a purely lab-based story.
Behind the hardware, Nokia is building an AI RAN ecosystem. At Mobile World Congress 2025, it announced an AI RAN center in Dallas that will enable partners like KDDI, SoftBank, and T-Mobile to develop and test AI-powered radio networks under realistic conditions, with the goal of shaping a platform-as-a-service model for operators. The idea is that carriers will eventually be able to host AI workloads at the edge of their networks, using Nokia's anyRAN architecture to share compute between radio and AI applications, cutting costs and opening new revenue streams.
For me, this is where the insider buying starts to make sense. Nokia is positioning itself as the glue between mobile networks and the AI infrastructure that Nvidia is building. It is not trying to compete with Nvidia's GPUs or large language models. Instead, it is trying to become the default way those models reach phones, cars, and factories over 5G and 6G. If that strategy works, Nokia's AI story will be less about selling boxes and more about selling intelligent, programmable network platforms.
