The French automation giant's acquisition of Oslo-founded Cognite, the largest Norwegian software exit on record, signals that the race to own the industrial AI stack is entering a consolidation phase where incumbents buy rather than build.
There's a specific kind of acquisition that changes a market's geometry rather than just its scoreboard. Schneider Electric's agreement to buy Cognite for $3.1 billion is that kind. Announced in late June 2026, the deal gives Schneider ownership of the industrial data and AI platform that connects raw operational technology from oil rigs, factories, and energy grids to the analytics and AI layers that actually make those assets perform better. That sounds like infrastructure plumbing, and it is. But control of the plumbing is precisely the point.
Cognite was co-founded in Oslo in 2016 by John Markus Lervik, Geir Engdahl, and Stein Danielsen, spun out of Aker, the Norwegian industrial conglomerate that held a major stake and is now realizing roughly 20 times its invested capital in the deal. Aker's estimated cash proceeds come to $1.48 billion. Saudi Aramco, Accel, and TCV were also shareholders. The company built its reputation on an industrial data platform called Cognite Data Fusion, which ingests messy, heterogeneous operational data from legacy equipment and makes it legible to AI systems running predictive maintenance, digital twins, and process optimization. Its customers include energy majors and industrial operators across pharmaceuticals, manufacturing, and infrastructure.
For Schneider, which already competes directly with Siemens and Honeywell in energy management and industrial automation, Cognite solves a specific strategic problem. You can sell a factory operator sensors, switchgear, and automation software, but if you don't own the data contextualization layer, you're dependent on a third party every time a customer wants to build an AI-driven workflow on top of your hardware. That dependency is fine when AI is a feature. It's untenable when AI becomes the core product. According to analysis from the ARC Advisory Group, Schneider ranked first on implementation criteria among industrial software vendors but lagged Siemens on generative AI depth. Cognite is the acquisition that addresses that gap directly.
The consumer AI acquisition frenzy of the early 2020s was mostly about talent and model capabilities. What's happening in industrial AI is structurally different. PwC's 2026 mid-year industrials deal outlook put total industrial manufacturing M&A at $173 billion over the past year, a 28% increase over the prior year, with mega-deals above $5 billion now accounting for 56% of deal value, up from 18% in fiscal year 2024. CB Insights counted 266 AI M&A deals in Q1 2026 alone, a 90% year-over-year increase. Eaton and Rockwell Automation are acquiring robotics startups. Hyperscalers are chasing data center cooling and power infrastructure. Everyone is buying the picks-and-shovels layer rather than competing on model benchmarks.
The reason is simple: industrial AI value doesn't live in the model, it lives in the data. A large language model with no access to a refinery's actual sensor history is useless to a plant manager. Cognite's value proposition was always that it could take decades of siloed operational data and make it queryable, contextual, and AI-ready. That is not a capability Schneider could replicate quickly in-house, and at $3.1 billion, it wasn't priced as if they could. The multiple reflects scarcity, not just revenue.
Frankly, the deal also tells you something about what Siemens and Honeywell will be doing in the next eighteen months. When a competitor acquires the market's most credible industrial data platform, the remaining players either build aggressively or go shopping. Siemens has Mindsphere and its own industrial IoT stack. Honeywell has Forge. Neither is as clean a third-party platform as Cognite, which deliberately served customers across competing automation ecosystems. That neutrality is now Schneider's competitive asset to manage, or risk losing the customer trust that made Cognite worth buying.
What the exit means for Oslo
The $3.1 billion headline is also a Nordic data point worth reading carefully. Cognite became Norway's first software unicorn after a $150 million Series B in 2021 valued it at $1.6 billion. The final exit at $3.1 billion is nearly double that figure, and the Aker release describes it as the largest Norwegian software and AI exit to date, ranking among the largest industrial software transactions in Europe. Cognite had already relocated its global headquarters from Oslo to Tempe, Arizona in early 2025, a move that says something honest about where enterprise software companies feel they need to be to close the deals that lead to exits like this one. The technology was built in Oslo. The exit was structured for a global market on American terms.
That tension is real but shouldn't obscure what the deal confirms. The Nordic deep-tech ecosystem, built on decades of oil and gas digitalization expertise, produced a company valuable enough for a French industrial giant to pay a 20-times return on invested capital to own. The institutional knowledge embedded in Cognite about how to make operational technology data usable is not something you grow in a hyperscaler's R&D lab. It comes from years of solving the specific, unglamorous problems of data integration in environments where sensors weren't designed to talk to each other and data governance was an afterthought. That combination of domain depth and platform architecture is exactly what makes industrial AI hard to commoditize, and exactly why Schneider Electric decided to buy it rather than build it.
The deal is expected to close in the coming quarters, pending regulatory approvals. When it does, the industrial AI market will have one fewer independent platform and one more reason for every automation incumbent to revisit its own data-layer strategy before the window closes.
Also read: X launches an official MCP server and every social platform will need to follow • Anthropic's refusal to bend to Washington has cost it Pentagon contracts and earned it a court fight it did not expect • Jane Street and TSMC's venture arm have backed Etched, the startup betting that locking silicon to transformers beats Nvidia on inference economics
