IAEA Outlook: Fusion Investment Accelerates but $80bn More May Be Needed for Commercialization

The International Atomic Energy Agency, the U.N. nuclear watchdog, said in its new World Fusion Outlook 2026 that investment in fusion energy is accelerating, but getting the technology to commercial use will still require a far bigger financial push. The report, released Thursday in Vienna, also says artificial intelligence could help speed progress — if the industry can solve a major problem with access to usable data.

The fourth edition of the outlook was launched at the Third High-level Meeting of the IAEA World Fusion Energy Group. In the report’s financing contribution, the investment advisory firm FusionX says the industry may need much more capital beyond what has already been raised. As the IAEA summary puts it, “FusionX estimates that a further US $80 billion in private capital could be required over the next 20 years for the industry to reach commercialization.” The same IAEA summary says private-sector fusion has raised $21.9 billion in committed capital, with more than 90% of that arriving in the past five years. It says private fusion companies attracted $4.5 billion in new funding in the 12 months leading up to July 2026, and that annual private funding this year could reach a record $7.6 billion if the current pace continues.

Those numbers matter because the sector is starting to look more like an emerging industrial buildout than a purely scientific effort. The report says private fusion companies spent more than $530 million with suppliers in 2025, a sign of growing activity across manufacturing and supply chains. It also lists 26 planned fusion power plants, with announced target dates ranging from the early 2030s to the 2050s. Separately, the IAEA said its Fusion Facility Database now records 174 fusion facilities worldwide — operating, under construction or planned — and that privately funded fusion facilities have tripled since 2021. The agency also said ITER, the large international fusion project in France, is executing ahead of its 2024 baseline schedule.

On AI, the report says the technology could accelerate fusion research and engineering through faster simulations, improved experiments and real-time control, better use of experimental data, and enabling technologies. But it identifies a major bottleneck: fusion data is fragmented and often difficult to access across institutions and projects. The report says the IAEA’s Fusion Data Lake is one attempt to address that problem. The IAEA summary warns that “Building the right data foundations could prove as important as advances in AI itself.” A contribution referenced from Google DeepMind’s April 2026 analysis makes a similar point, saying, “This demands what we’re calling AI data stocktakes.”

The outlook also looks beyond first-of-a-kind plants to the harder question of scale. In deployment scenarios contributed by EUROfusion, a European fusion consortium, the eventual role of fusion in the global power system depends less on the exact date the first plant arrives than on how quickly and how widely plants are built after that. Under different assumptions, the report says fusion could supply anywhere from a few percent to 15% to 20% of global electricity by the end of the century.

That framing is important for a technology that remains in development. The IAEA’s report presents fusion as a field with growing investor interest and a widening industrial base, but also one whose commercial impact will depend on financing, supply chains, regulation and deployment speed as much as on technical breakthroughs.

Funding estimates, however, depend heavily on methodology and time horizon. Research cited in the report notes that FusionX used different assumptions in an earlier 2025 paper and reached different top-line numbers, so the $80 billion figure should be understood as this report’s attributed estimate, not a universal consensus number.