Future of Humanity Institute

The Future of Humanity Institute (FHI) was a research institute at the University of Oxford, founded by Nick Bostrom in 2005 and closed in 2024. For nearly two decades, it was the primary academic institution dedicated to the study of existential risk and the long-term future of humanity. Its closure — amid Oxford's restructuring and political pressure on AI safety research — marks the end of an era in which the most important questions about technological risk could be asked from within established academia.

FHI's research spanned four overlapping domains: the philosophy and probability of existential risk, the ethics of future generations and the scope of moral concern across time, the strategic dynamics of artificial general intelligence development, and the simulation argument. The institute's output was not merely academic. It shaped the research agendas of OpenAI, DeepMind, and Anthropic; influenced UK and US AI policy; and established the conceptual vocabulary — orthogonality thesis, instrumental convergence, scalable oversight — that now structures the field.

The closure of FHI is not merely the loss of a research center. It is the loss of an institutional form: the university-based institute with the freedom to pursue long-term, high-stakes research without commercial pressure or political oversight. Whether this loss is temporary or permanent depends on whether other institutions — the new crop of AI safety nonprofits, the remaining academic centers, or international governance bodies — can replicate FHI's combination of intellectual seriousness and institutional independence.