Clean Energy & AI: 3 Key Takeaways from the 2026 Norrsken100 List

The 2026 Norrsken100 list of global impact startups is led by two categories: clean energy and electrification (26% of companies) and artificial intelligence (25%). The takeaway for the energy sector is clear: the next phase of the transition depends as much on intelligent software that runs existing assets as on building new ones.

What is the Norrsken100, and why does the 2026 list matter?

The Norrsken100 (formerly Impact/100) is an annual list compiled by the Norrsken Foundation, the Stockholm-based nonprofit that backs entrepreneurs working on the world’s biggest challenges. For the 2026 edition, Norrsken processed more than 1,400 nominations from over 80 partners to select 100 companies with the potential to “fix the future.” This year’s cohort was celebrated on the Nasdaq Tower in Times Square.

The list is a useful signal of where impact capital and founder talent are heading. In 2026, that signal points in one direction: clean energy and AI are converging. Together they account for roughly half of the companies selected.

Delfos is one of the 100 companies selected for the 2026 Norrsken100. We read this as confirmation of something we see every day in operations. The energy transition is no longer only a hardware problem. The wind, solar and storage capacity being built today has to be operated, diagnosed and optimized for 25 years or more, and that is where AI makes the biggest difference.

Takeaway 1: Clean energy and electrification is the largest sector on the list

Clean energy and electrification is the single largest category in the 2026 Norrsken100, representing 26% of the selected companies.

The reason is scale. Wind, solar and battery storage are coming online globally at a pace that existing infrastructure, workflows and software were never designed for. Every new gigawatt creates demand for better tools across the full asset lifecycle, and the 2026 list reflects that breadth:

What these companies share is a focus on making clean energy work better at scale, not just on producing more of it. Delfos works in the same space. Since 2017 it has built AI-driven performance tools for wind, solar and storage, and it now supports more than 1,000 energy sites across more than 10 countries.

Takeaway 2: AI has moved from hype to operational reality

A quarter of the 2026 Norrsken100 companies (25%) put artificial intelligence at the core of their impact model. In energy, that shift is from AI as a promise to AI as part of daily operations.

Building a renewable plant is only the first step. Once assets are commissioned, portfolios tend to grow much faster than the engineering teams responsible for running them. A performance team that managed 500 MW a few years ago may now be responsible for several gigawatts across wind, solar and storage, often in multiple countries.

General-purpose AI chatbots don’t solve this. Critical energy infrastructure needs AI that understands turbine behavior, inverter faults, SCADA signals and the specific diagnostic logic engineers use to find root causes.

That is the idea behind vertical, domain-specific AI, and it is what Delfos built with its AI Virtual Performance Engineer. It learns from more than nine years of historical diagnostic workflows, so instead of handing site teams a stream of raw alarms, it delivers prioritized, actionable root-cause insights: what is underperforming, why, and what to do about it.

“The energy transition will only succeed if existing infrastructure runs far more efficiently and reliably than it does today.”

Guilherme Studart, CEO and Co-founder, Delfos Energy

Takeaway 3: AI lets O&M teams scale without burning out

The most important effect of operational AI in renewables is that it lets small engineering teams manage large portfolios without being overwhelmed. It augments engineers; it does not replace them.

Consider the daily reality of a performance engineer. A multi-gigawatt portfolio can generate thousands of SCADA alarms every day. No one sitting at a desk can investigate each one, so teams triage by instinct, important issues get missed, and repetitive diagnostic work crowds out the analysis that actually improves asset performance.

Operational AI changes that split. By automating the repetitive first pass (filtering noise, correlating signals, identifying likely root causes), it frees engineers to focus on high-value interventions: planning maintenance, negotiating with OEMs, and making decisions that protect revenue.

In February 2026, Delfos announced a partnership with Dost Enerji to deploy its AI Virtual Engineer across a wind and solar portfolio in Turkey. For a lean team managing complex wind assets, the previous way of working was no longer sustainable. That is the gap operational AI is built to close.

The long-term payoff is measured in lifecycle performance. Wind, solar and storage assets are designed to run for 25 years or more. Small losses that go undetected, such as a misaligned yaw system or a degrading string of panels, compound over that lifetime. AI that catches them early helps every site deliver what it was built to deliver.

Conclusion: the next era of clean energy runs on intelligent software

The 2026 Norrsken100 list makes a clear case. Clean energy remains the largest area of impact innovation, and AI is the force that will determine how efficiently that energy is produced, managed and maintained.

Delfos is proud to be among the 100 companies on the 2026 Norrsken100. Our AI Virtual Performance Engineer gives O&M teams a way to see which assets are losing energy, why, and what to fix first, so the same team can cover a growing portfolio.

Explore the full 2026 list at norrsken.org/100.