Calling on policymakers to implement a global moratorium on large AI training runs until alignment is solved.
Calling on policymakers to implement a global moratorium on large AI training runs until alignment is solved.
People– no linked people
Updated 04/02/26Funding Details
Updated 04/02/26- Annual Budget
- -
- Current Runway
- -
- Funding Goal
- -
- Funding Raised to Date
- -
Org Details
Updated 04/02/26Global AI Moratorium (GAIM) is an advocacy campaign operated by the AI Safety and Governance Fund (AISGF), a 501(c)(4) nonprofit dedicated to ensuring AI development is safe, aligned with human values, and does not pose an existential threat to humanity. The campaign's primary public face is the website moratorium.ai, which marshals evidence from leading AI researchers, public opinion surveys, and technical analysis to argue that a global moratorium on frontier general AI training runs is necessary to prevent human extinction. The campaign's central argument rests on three pillars: that leading AI laboratories are racing to build systems smarter than humans within a decade; that no one currently knows how to make such systems reliably safe or aligned with human values; and that once a misaligned superintelligent system exists, it will be impossible to control. The site quotes Nobel laureates, AI lab CEOs, and public officials including Geoffrey Hinton, Demis Hassabis, Yoshua Bengio, Sam Altman, and UN Secretary-General António Guterres to illustrate the breadth of expert concern. The campaign calls for concrete policy steps: imposing strict oversight requirements on large-scale AI training runs, building international mechanisms to monitor and restrict access to the compute infrastructure needed for frontier models, and ultimately enforcing a global halt on general-purpose AI development until alignment research produces a verified solution. The site emphasizes that restrictions on frontier training runs would not significantly harm beneficial narrow AI applications in medicine, science, or industry. AISGF also operates a secondary site (aisgf.us) that hosts an AI-safety chatbot and promotes the book If Anyone Builds It, Everyone Dies by Eliezer Yudkowsky and Nate Soares. The organization currently relies on volunteer and unpaid staff and is seeking donations to scale its communications and advertising efforts. No specific founders, leadership names, team size, or budget figures are publicly disclosed.
Theory of Change
Updated 04/02/26GAIM believes the primary lever for reducing existential risk from AI is political will: if enough policymakers, opinion leaders, and members of the public understand the alignment problem and the extinction-level stakes, governments can coordinate an international moratorium on frontier AI training before misaligned superintelligence is created. The causal chain is: public education and persuasion campaigns raise awareness of AI risk, increased public concern creates political pressure on legislators and regulators, governments negotiate and enforce compute-governance agreements and training-run restrictions, and the resulting pause buys time for alignment research to mature. The campaign specifically targets the compute layer as the most tractable enforcement point, arguing that monitoring chip production and data-center access makes a moratorium verifiable and enforceable even without perfect global cooperation.
Grants Received– no grants recorded
Updated 04/02/26Projects– no linked projects
Updated 04/02/26Discussion
No comments yet. Be the first to share your thoughts.