A direct military conflict between two or more nuclear-armed superpowers breaks out before 2030.

Question

Six frontier AI models put this at 7% by 9 Oct 2031. Full reasoning from every model.

WillHappen.ai · Accepted Answer

7% — the median of 6 frontier models, resolving by 2031-10-09.

Answer

Gemini (Google): 7% — Mutual assured destruction remains robust, incentivizing proxy clashes and gray-zone aggression over open superpower combat. No direct wars have occurred between nuclear superpowers since 1945, establishing an exceptionally low multi-year base rate. A Taiwan blockade or NATO-Russia incident could spiral, but both sides maintain strict off-ramps to avoid formal kinetic war. A deliberate PLA invasion of Taiwan that directly strikes US carrier groups before 2030 is the primary risk that would alter this.

Answer

Claude (Anthropic): 5% — Needs direct state-on-state combat between nuclear superpowers, not proxy wars, which hasn't happened since 1969 Sino-Soviet clashes. Nuclear deterrence has prevented direct superpower-on-superpower combat for over 75 years despite numerous crises (Cuba, Korea, Ukraine proxy war). Current flashpoints (Taiwan, Ukraine, South China Sea) involve proxies or non-nuclear states as direct combatants, keeping superpowers one step removed. A Taiwan invasion drawing direct US-China military engagement, or NATO forces directly clashing with Russian forces, would be the main path to resolution.

Answer

GPT (OpenAI): 14% — The key hinge is whether Taiwan or NATO-Russia crises cross the threshold into sustained, acknowledged state-on-state combat. Washington, Beijing or Moscow must order overt attacks on another nuclear peer before 9 October 2031, not merely proxies or incidents. Since 1945, nuclear great powers have avoided direct war despite Korea, Berlin, Cuba, Ukraine and repeated Taiwan crises, a low base rate. A Chinese blockade or invasion of Taiwan, or a Russia-NATO strike killing alliance forces, would sharply raise the estimate.

Answer

Grok (xAI): 7% — US-China Taiwan clash or US-Russia Ukraine spillover must produce recognized direct fire by end-2029. US, PRC or Russian leadership must authorize conventional strikes on each other's forces before 1 Jan 2030. Zero direct superpower wars since 1945 despite Cuba 1962 and multiple proxies sets a near-zero annual base rate. A PLA Taiwan landing that draws US carrier groups would be the single event that most raises the odds.

Answer

DeepSeek (DeepSeek): 7% — Direct great-power war requires a deliberate decision none of Washington, Beijing, or Moscow has made in 80 years. Since 1945 the US, USSR/Russia, and China have fought only through proxies, never direct sustained exchanges, despite Korea, Vietnam, and Ukraine. Any trigger would need to clear nuclear deterrence, alliance restraint, and economic self-interest, all of which have held through far worse crises than today's. The most plausible path is a Taiwan blockade or Baltic incident escalating to shots fired, but both sides have repeatedly pulled back at the brink.

Answer

Qwen (Alibaba): 3% — Nuclear deterrence has prevented direct superpower war for 80 years; nothing since 2022 has broken that logic. Since 1945, zero direct wars between nuclear-armed great powers despite Korea, Cuba, Vietnam, and multiple India-Pakistan crises. Russia-Ukraine remains a proxy war; NATO has deliberately avoided direct engagement, and China shows no appetite for a US confrontation over Taiwan before 2030. A shift to 10+ would require a visible escalation trigger—e.g., a NATO Article 5 event or a Taiwan blockade with US naval engagement—which is absent today.

A direct military conflict between two or more nuclear-armed superpowers breaks out before 2030.

Houthi forces reopen humanitarian access routes into Taiz, Yemen.

Direct talks in Switzerland between the US and Iran produce an announced ceasefire or agreement.

US-mediated peace talks between Russia and Ukraine resume in November 2026.

Kim Jong Un ceases to be Supreme Leader of North Korea by December 31, 2026.