Overview

• The proposed checklist outlines ten governance priorities for AGI, including alignment and safety, regulation, IP and access, economic impacts, national security, ethical use, transparency, control and off‑switch design, public trust, and existential‑risk management.
• The framework is intended for use across three phases—Pre‑AGI, Attained‑AGI, and Post‑AGI—to steer policy and engineering decisions as capabilities evolve.
• Coverage reiterates that AGI has not been attained and that timelines remain uncertain, with forecasts ranging from decades to centuries and lacking solid evidence.
• Existing resources from the United Nations and NIST are cited as foundations for risk management, yet reporting notes a compliance gap and emphasizes that laws exert stronger pressure than voluntary ethics.
• Definitional disagreement persists, with references to OpenAI’s Charter tying AGI to outperforming humans at most economically valuable work and to Sam Altman’s recent remark that the term is “not a super useful term.”