President Trump’s September 22, 2026 address to the United Nations General Assembly
The developing sequence itself may be significant: artificial intelligence → generative AI → frontier AI → agentic AI → capability thresholds → superintelligence. Formal standards organizations are standardizing much of the first half of this vocabulary while laboratories are still creating much of the second. There is not yet a mature standards catalog specifically for superintelligence. Instead, a vocabulary and governance architecture is forming around frontier AI, general-purpose AI, capability thresholds, evaluation, scaling and catastrophic risk.
ISO/IEC 22989 — Artificial intelligence concepts and terminology
The foundational vocabulary standard for artificial intelligence. It establishes common concepts and terms — a natural Language Monday starting point because a new technical order first requires a stable language.
ISO/IEC 42001:2023 — Artificial intelligence — Management system
The international AI management-system standard. It applies policies, responsibilities, risk controls and continual improvement to organizations developing or using AI.
ISO/IEC 23894:2023 — Artificial intelligence — Guidance on risk management
Guidance for identifying and managing AI-related risks throughout organizational activities. Its technology-neutral approach makes it relevant as AI capabilities increase.
ISO/IEC 42005:2025 — AI system impact assessment
A newer part of the standards architecture providing a structured approach to assessing the consequences associated with AI systems.
NIST Artificial Intelligence Risk Management Framework (AI RMF)
A voluntary U.S. framework organized around four functions: Govern, Map, Measure and Manage. NIST is developing the framework alongside guidance addressing generative AI.
NIST AI 200-2 — TEVV-Athlon Framework for Evaluating AI Systems
An emerging framework for Testing, Evaluation, Verification and Validation (TEVV) of AI systems, including large language, multimodal and agentic systems.
Responsible Scaling Policy (RSP)
An emergent private specification rather than a consensus standard. It links increasingly capable AI models with progressively stronger safety, evaluation and security requirements.
“Composition A” 1923 Piet Mondrian