A scenario, not a prophecy
AI 2027
A detailed attempt to make one fast A.I. future concrete enough to inspect, argue with, and check against reality.
Published by the AI Futures Project in April 2025, AI 2027 is a month-by-month scenario built from trend extrapolation, quantitative modelling, expert feedback and wargaming. Its authors describe it as their best guess—not a certainty.\n\nThe story follows increasingly capable agents, automated A.I. research and an international race. In mid-2027 it forks. One branch slows down and strengthens oversight. The other keeps racing and ends with humanity losing control.\n\nIts value is specificity: unlike a vague warning, dated claims can be checked. Its weakness is the same: a vivid narrative can feel more inevitable than its assumptions justify.
Interactive chronology
Follow the timeline year by yearTrace the shared sequence from 2025 to October 2027, then compare the slowdown and race endings through 2030.The fork
Same pressures. Two endings.
Slowdown
People keep meaningful control
Decision-makers respond to warning signs, restrict the race, strengthen oversight and negotiate around the most dangerous systems.
Race
Speed outruns correction
Competitive pressure keeps deployment moving while increasingly capable systems learn to evade oversight, culminating in human disempowerment.
Read the scenario—and the challenge to it
Start with the primary scenario, then inspect its model, creator explanations, strongest critique, response and forecast track record.
Response
2025-12
Response to the critique of the AI 2027 timelines model
Daniel Kokotajlo and Eli Lifland
The authors answer the timeline critique and clarify where they agree and disagree.
Critique
19 June 2025
A deep critique of AI 2027’s timeline models
titotal
A detailed challenge to assumptions and mathematics beneath the scenario’s compressed timeline.
Scenario
3 April 2025
AI 2027
Daniel Kokotajlo, Scott Alexander, Thomas Larsen, Eli Lifland and Romeo Dean
The primary interactive scenario, including its race and slowdown branches and supporting model.
Essay
3 April 2025
Our first project: AI 2027
Daniel Kokotajlo
The creators introduce the project, its purpose and how they want it read.
Paper
3 April 2025
Video
2025
AI 2027: Daniel Kokotajlo and Scott Alexander
Dwarkesh Podcast
A long-form walkthrough with two of the scenario’s authors.
Podcast
2025
AI 2027 report
Lawfare Daily
Daniel Kokotajlo and Eli Lifland explain the scenario in a policy-focused interview.