Every year, the Forum selects 100 early-stage companies to join its Technology Pioneers Community. This year’s cohort, spanning 23 countries, arrives at a defining moment: deep tech investment is surging, AI accounts for more than 60% of all global venture capital, and the lines between software, hardware, energy and biology are increasingly blurred.

The chosen companies are tackling problems from nuclear fusion and lunar resource extraction to neurosurgical robotics and post-quantum cryptography. Together, they reflect five structural shifts in how frontier technology is being built and funded.

“What strikes me most is that they’re working on challenges that until recently would have required enormous budgets, infrastructure and headcount," said Verena Kuhn, Head of Innovator Communities at the World Economic Forum.

And AI, she adds, is not just the subject of this cohort but also the reason it is possible.

1. The arrival of an AI infrastructure layer

The AI investment journey, which began with money flowing into application-layer companies (think chatbots and vertical wrappers sitting on foundation models), is giving way to something more structurally consequential: rising funding for start-ups building the infrastructure to support an agent economy.

These start-ups are forming with notable speed, signalling where value can be created in AI. Eight such companies from the 2026 cohort sit in this category – none of which existed as commercial products three years ago.

These include Skyfire, which provides verified identity and payments for AI agents; Paid, which handles billing, pricing and subscription management for agent services; VESSL AI, which orchestrates GPU workloads at scale; Adaption, which eliminates the need for expensive retraining cycles with AI that continuously learns; Inception, which builds diffusion-based large language models that the company says run several times faster than pattern-recognizing autoregressive architectures; and Odyssey, which builds world models for agents that learn through simulation to predict and interact with the world.

Underpinning that agent economy, however, is a physical infrastructure challenge that several pioneers are tackling head-on.

2. The convergence of energy and compute

Data centres are expected to consume twice as much power by 2030 due to surging AI demand. Eight companies in the 2026 cohort are addressing the underlying challenge, and the breadth of their approaches – from geothermal and fusion to space-based solar – show there is currently no dominant solution.

Emerald AI and GridCARE are addressing grid orchestration and capacity forecasting for data-centre developers – helping facilities get online by unlocking ‘hidden’ capacity on existing grids; IONATE is building intelligent hybrid transformers to replace ageing grid hardware and optimize for efficiency and resilience at a system level; Overview Energy proposes beaming space-based solar via satellite to Earth's grids; Realta Fusion and Mazama Energy are targeting clean power generation at industrial scale, in the spaces of compact, modular fusion systems and geothermal energy that extracts heat from subsurface “superhot” rocks respectively; and Power to Hydrogen and Pure Lithium bring new approaches to energy storage.

3. The emergence of thematic tech clusters

Beyond the AI infrastructure and energy convergence themes above, the 2026 cohort clusters into six further areas of frontier technology:

4. Deep tech’s spread beyond Silicon Valley

The growing role of India, China and Europe in frontier technology became clear with the 2025 Technology Pioneers cohort, and that trend deepens this year.

Five of this year’s companies are working in AI and quantum computing in South Korea.

In the Middle East, Saudi Arabia’s Intella is addressing a critically underserved language technology gap with Arabic-first speech AI for 400 million speakers.

And in Colombia, Quipu is addressing financial exclusion with AI credit assessment, helping the informal workforce to access conventional financial services.

5. The prominence of two founder archetypes

The final theme emerging from the companies in the 2026 list focuses on the founders: throughout the cohort, there is a high concentration of two archetypes.

The first is those with decades in the laboratory under their belt – several cohort companies are led by scientist-founders or researchers who have built long careers in the lab before commercializing their work.

The second is the repeat builder, infrastructure founders who have already exited the layer beneath their current company. Examples include GridCARE’s Amit Narayan, who built a software solution for virtual power plants and distributed energy resources before tackling grid AI; and RLWRLD’s Junghee Ryu, who started a computer vision technology company, Olaworks (acquired by Intel), before building robotic foundation models.

Taken together, these five themes suggest innovation in frontier tech is increasingly systems-based and shaped by not just ideas but factors including access to power, compute, supply chains and more.

Translating breakthroughs into real results at scale, and across industries and economies, will require partnerships and collaboration to pool capabilities and knowledge, share risk and build supporting ecosystems.

The Forum is spotlighting how innovation moves from breakthrough to scale to impact ahead of 'Summer Davos' in China, 23–25 June 2026. Follow the latest.