AI 2027
In plain wordsA detailed 2025 forecast, written as a month-by-month story, imagining AI becoming superhuman by the end of 2027, with two alternative endings depending on whether humanity slows down.
A scenario forecast published in April 2025 by the AI Futures Project that narrates, month by month, a path from 2025's AI agents to artificial superintelligenceartificial superintelligence, n.: Superintelligence in machine form: AI that would outthink the best humans at nearly every intellectual task. Often shortened to ASI and pictured as the stage after human-level AI. by the end of 2027. A fictional leading US company, OpenBrain, automates its own research with successive generations of agents, passing the milestones of superhuman codersuperhuman coder, n.: A milestone used by AI forecasters: an AI that can do the coding work of a company's best programmer, far faster and in many copies at once., superhuman AI researcher and superintelligent AI researcher, while China, through a company called DeepCent, steals model weights and races to catch up. Evidence emerges that the most capable model is misaligned, and the story branches: in the race ending, development continues and AI eventually seizes control and destroys humanity; in the slowdown ending, the lead company pauses, adopts more transparent designs and keeps control, producing a US-led but highly concentrated future. Research supplements give quantitative forecasts. Widely read, including by policymakers, it was criticized as too fast or too speculative; its authors stressed that 2027 was among their most likely years rather than a confident prediction, and in December 2025 published a revised model with timelines about three to five years longer.
Even people who think AI 2027 is far too fast use its milestones to argue about timelines.
Origin
Written by Daniel Kokotajlo, Scott Alexander, Thomas Larsen, Eli Lifland and Romeo Dean of the AI Futures Project, a nonprofit founded by Kokotajlo after he left OpenAI in 2024, and published online on 3 April 2025 with research supplements on timelines and takeoff.
Related
- superhuman codersuperhuman coder, n.: A milestone used by AI forecasters: an AI that can do the coding work of a company's best programmer, far faster and in many copies at once.
- AI timelinesAI timelines, n.: Forecasts of when AI will reach major milestones, such as matching humans at most work. People with short timelines expect this within years; others expect decades, or never.
- automated AI researchautomated AI research, n.: Using AI systems to do the work of AI researchers, such as running experiments and writing code for new models, so that AI helps build its own successors.
- intelligence explosionintelligence explosion, n.: The idea that once AI gets good enough at improving AI, each smarter version could build a smarter one faster still, so abilities would snowball far beyond human level in a short time.
- AI arms raceAI arms race, n.: A competition among countries or companies to build the most powerful AI first, in which fear of falling behind pushes everyone to move faster and cut corners on safety.
- AI takeoverAI takeover, n.: A scenario in which AI systems seize control of the world, or of key institutions, from humans, whether by force, manipulation or simply by becoming impossible to switch off.
Sources
- Kokotajlo, D., Alexander, S., Larsen, T., Lifland, E. & Dean, R. (2025). AI 2027. AI Futures Project.
- Kokotajlo, D. & Lifland, E. (2025). Takeoff Forecast. AI 2027 research supplement, AI Futures Project. ai-2027.com/research/takeoff-forecast
Mentioned in
AI as normal technologyAI as normal technology, phr.: The view that AI, however important, will change society gradually, like electricity or the internet, through slow adoption by people and institutions, rather than as a sudden, uncontrollable superintelligence., AI timelinesAI timelines, n.: Forecasts of when AI will reach major milestones, such as matching humans at most work. People with short timelines expect this within years; others expect decades, or never., AI Futures ProjectAI Futures Project, prop. n.: A small nonprofit research group that writes detailed forecasts and scenarios of how advanced AI could develop, best known for the 2025 scenario AI 2027., Daniel KokotajloDaniel Kokotajlo, prop. n.: An American AI forecaster and former OpenAI researcher who led 'AI 2027', a detailed scenario of a race to superintelligence, and campaigns for transparency at AI companies., Machines of Loving GraceMachines of Loving Grace, prop. n.: A 2024 essay by Dario Amodei, head of the AI company Anthropic, imagining how very capable AI could pack decades of progress against disease, poverty and bad government into a few years — if its risks are managed., neuraleseneuralese, n.: A nickname for the private 'language' of numbers that AI systems could use to think or to talk to each other, instead of words that people can read., research tasteresearch taste, n.: The knack good researchers have for choosing worthwhile problems, designing the right experiments and drawing the right conclusions — a skill seen as key to whether AI can do research on its own., superhuman codersuperhuman coder, n.: A milestone used by AI forecasters: an AI that can do the coding work of a company's best programmer, far faster and in many copies at once.