Future of Humanity Institute: AI Risk Research, Closure

AI risk inside a university institute

The Future of Humanity Institute was a multidisciplinary group within Oxford's Faculty of Philosophy, founded in 2005. Its research addressed long-term risks from technology, including advanced AI, alongside broader questions about humanity's future. Oxford Martin School support ended in 2012, so that historical funding relationship should not be read as its final organizational home. [1][2]

Its outputs included Nick Bostrom's Superintelligence and work on research priorities for robust and beneficial AI. These are contributions to analysis and research agendas, rather than shipped AI products or evidence that every later safety institution descended from FHI. [2]

FHI closed on April 16, 2024. Its closing statement describes a hiring and fundraising freeze beginning in 2020 and the Faculty's late-2023 decision not to renew remaining contracts. Those explanations are the institute's own account; an archived university program page using present tense does not establish that it still operates. [1][2]