Rank #12 ITN Score: 19/30

Great Power Conflict

Military competition between major powers, especially around AI and nuclear weapons, raises the risk of catastrophic war. AI-accelerated decision cycles and autonomous weapons lower the threshold for escalation.

Key Statistic

Nuclear weapons states possess ~12,500 warheads; AI could compress decision timelines from days to seconds.

ITN Assessment

Importance9/10

Scale & severity

Tractability4/10

Ease of progress

Neglectedness6/10

Lack of attention

0 followers

Why This Matters

Existential risk potential; low tractability for non-state actors; receives some attention but AI-specific dimensions are underexplored.

Key Organizations

Nuclear Threat Initiative
Reducing threats from nuclear, biological, and other weapons.
Ploughshares Fund
Grantmaker focused on reducing nuclear dangers.
Carnegie Endowment for International Peace
Global policy research including nuclear and great-power issues.
Federation of American Scientists
Research on nuclear arsenals, AI, and global security.
Council on Strategic Risks
Analyzes catastrophic risks including great-power conflict.

Submit New Draft

1 approved

Sort

Nuclear Shadows and Digital Accelerants: AI's Role in Great Power Competitionv1Accepted

WorldProblems Consortium

Military competition between major powers has always carried catastrophic risk. AI introduces new accelerants: compressed decision cycles, autonomous weapons that reduce human judgment in the kill chain, and dual-use capabilities that blur civilian and military targets. This report examines how AI changes the risk calculus of great power conflict, evaluates existing arms control frameworks, and proposes a research and diplomatic agenda for AI-specific risk reduction.

+0 votes

Followers

Approved Docs

Evidence & References

80,000 Hours Problem Profile

https://80000hours.org/problem-profiles/great-power-conflict/

Problem Profile

Why Pressing

US-China strategic competition, Russia's war in Ukraine, and AI-enabled military modernization raise the probability of a major-power conflict. Global nuclear arsenals still total roughly 12,000 warheads, and arms-control treaties have eroded. A US-China war over Taiwan or escalation in Europe could credibly risk nuclear use. AI in early-warning, command-and-control, and autonomous weapons introduces new escalation pathways that have not been stress-tested.

Scale

Direct deaths from a US-Russia nuclear exchange are estimated in the hundreds of millions, with climatic and agricultural follow-on effects potentially killing several billion more. Conventional great-power war alone would be catastrophic. Toby Ord estimates roughly 0.1 percent existential risk from nuclear war this century, and meaningfully higher probability of non-existential catastrophe.

Neglectedness

Funding for nuclear-risk-focused work has shrunk dramatically since the 1990s. The Nuclear Threat Initiative operates on around 50 million dollars per year. The arms-control community is aging, and China-focused policy capacity in the US is limited relative to the strategic importance of the relationship.

Tractability

The track record includes SALT, START, New START, the INF Treaty, and various confidence-building measures. Current levers include stockpile reduction, launch-on-warning reform, keeping AI out of nuclear command-and-control, crisis hotlines, and track-II diplomacy. New arms-control agreements are politically hard now but may become feasible after crises.