A report commissioned by the U.S. State Department has warned of the existential threat from artificial intelligence. The report posited a worst-case-scenario "extinction-level threat" to humanity. The authors identified a "growing risk to national security posed by rapidly expanding AI capabilities from weaponisation and loss of control". The report added that there was a "clear and urgent need" for the U.S. government to intervene "quickly and decisively". It continued: "The rise of advanced AI has the potential to destabilise global security in ways reminiscent of the introduction of nuclear weapons." The report calls for a new federal agency to regulate companies developing AI tools and to limit the growth of AI.

The 247-page report was commissioned by the State Department in November 2022. It was written by a company that provides technical briefings and guidance on AI. The authors spoke extensively to experts at AI companies like OpenAI, Google DeepMind and Meta. The authors recommended setting an upper limit on how much computing power an AI model could have. They also recommended making it a criminal offence to open-source or reveal the coding behind more powerful AI models. The Futurism.com website stated: "Over half of the AI researchers surveyed say there's a five per cent chance that humans will be driven to extinction, among other 'extremely bad outcomes.'"