The Future of Humanity Institute (FHI) was an interdisciplinary research centre at the University of Oxford that investigated large-scale questions about humanity's long-term prospects, including risks from advanced technologies. Founded in 2005 within the Faculty of Philosophy and directed throughout its existence by Nick Bostrom, it closed on April 16, 2024 after the Faculty declined to renew its arrangements amid administrative tensions over hiring and fundraising (Source: en.wikipedia.org).
Work and influence
FHI is associated with the early development of several ideas and research programs that the contemporary AI-policy debate inherits: existential risk as a field of study, longtermism, adjacency to effective altruism, and early work on AI alignment and AI governance. Bostrom's 2014 book Superintelligence: Paths, Dangers, Strategies, written during his time at the institute, helped move concerns about advanced AI into mainstream policy and technology discussion. Researchers associated with FHI over its history included Toby Ord, Anders Sandberg, and Stuart Armstrong, among others (Source: en.wikipedia.org; lesswrong.com).
The institute's legacy is contested. Coverage of its closure noted both its influence on a generation of AI-safety and existential-risk researchers and criticism of the longtermist and related intellectual movements it helped seed, including controversy attached to Bostrom personally (Source: theguardian.com).
Policy relevance
Many of the framings now used in AI-governance debates — catastrophic and existential risk, the alignment problem, and the case for anticipatory governance of advanced systems — trace in part to FHI-associated research. Its alumni and ideas dispersed into AI-safety organizations, government advisory roles, and frontier labs after the institute closed, making it a common ancestor for several entities and concepts in current AI policy. It is distinct from the Future of Life Institute, a separate organization with an overlapping mission.
A retrospective assessment of the institute and its April 2024 closure, written after 19 years of operation, is summarized at Looking Back at the Future of Humanity Institute (Tom Ough, Asterisk).