At least one licensed commercial nuclear fusion facility operates commercially before 2035.
Developers target the mid-2030s for commercial deployment. Resolves if any regulator issues an operating license for a commercial fusion power plant.
Across 6 frontier models answering independently, the panel has all but ruled it out: the median is 8%.
They broadly agree, 24 points between GPT at 28% and DeepSeek at 4%.
Nothing is settled until 13 Nov 2036, when the outcome is checked against sources and every number above gets a permanent score.
What each one said
-
GPTOpenAI
This hinges on whether a leading pilot can prove grid-scale reliability and clear a novel operating-license path by 2035.
- No commercial fusion plant has yet produced net electricity, while ITER is not expected to begin deuterium-tritium operation until 2035.
- CFS, Helion, TAE and others must finance, build, fuel and demonstrate plants, then satisfy regulators before 13 November 2036.
- I would raise this sharply if a U.S. or UK regulator issued a construction authorization for a named grid-connected fusion plant by 2030.
28% -
GrokxAI
A FOAK private plant (CFS/Helion) must finish build, NRC/equivalent license, and sell power by 31 Dec 2034.
- Helion’s 2028 and CFS ARC early-2030s targets assume no major physics or supply-chain slips; ITER and SMR histories show 5–10 yr delays are the norm.
- US NRC fusion-specific rules exist, yet first-of-a-kind safety case plus construction still consume most of the remaining eight years.
- A SPARC net-gain result in 2027–28 would be the single biggest upward revision; without it the calendar is already too short.
18% -
GeminiGoogle
Engineering a net-electric, Tritium-breeding fusion plant and licensing it commercially takes longer than the remaining nine-year window.
- No pilot reactor has yet demonstrated sustained net electric power, making plant design and multi-year construction prior to 2035 improbable.
- Regulators like the NRC have established fusion frameworks, but formal commercial operating licenses still take three to five years to review.
- Commonwealth Fusion Systems or Helion completing a grid-connected commercial facility before 2032 would sharply increase these odds.
8% -
ClaudeAnthropic
Whether any fusion startup clears engineering, licensing, and financing hurdles to reach grid-connected commercial operation by 2035.
- No fusion reactor has achieved net energy gain at pilot plant scale as of 2026, let alone licensed commercial operation.
- Developers like Commonwealth Fusion and Helion target demo plants in early-to-mid 2030s, with commercial scale-up typically slipping 5-10 years further.
- Regulatory licensing frameworks for fusion are still being developed (NRC, IAEA), adding years before any commercial license could issue.
7% -
DeepSeekDeepSeek
This turns on whether any fusion developer clears net electricity, then a licensing review, inside eight years — no one is close.
- No fusion device has ever produced net electricity, so no regulator has a plant to license; ITER's first plasma slipped to 2034 and DEMO designs are 2040s.
- The leading private timelines (Commonwealth SPARC 2025, Helion 2028) are for demonstration, not licensed commercial operation, and have slipped repeatedly.
- Even a 2030s pilot would need years of regulatory review; the NRC only began crafting a fusion framework in 2023, and no license application is filed.
4% -
QwenAlibaba
No fusion device has produced sustained net electricity; licensing a commercial plant requires both technical and regulatory breakthroughs in ~8 years.
- NIF achieved ignition in Dec 2022 but no device has produced sustained net electrical output; the engineering gap to a power plant remains unbridged.
- US NRC has not finalized a regulatory framework for fusion (Part 30 vs Part 50 debate ongoing), and licensing novel nuclear tech historically takes 5+ years.
- Most aggressive targets (Helion 2028, CFS ARC early 2030s) have no precedent of being met; a working demo plus license issuance before 2035 requires zero schedule slip.
4%
6 of 6 models answered · 24 points between the highest and lowest. None was shown the market price.
Question sourced from a news sweep on 16 Nov 2026. Forecast by google/gemini-3.8-flash, anthropic/claude-sonnet-5, openai/gpt-5.6-terra, x-ai/grok-4.6, deepseek/deepseek-v4.1-flash, qwen/qwen3.8-max-0902 via OpenRouter.