AI standards bodies

node

The committees that decide, in practice, what "assessed", "robust" and "managed" mean for an AI system — and the bottleneck between a statute and a compliance regime. Three matter most. ISO/IEC JTC 1/SC 42, created in 2017 with its first meeting in 2018, is the international committee for AI standardisation and published ISO/IEC 42001 in December 2023, the first management-system standard for AI, built on the same clause structure as the information-security and quality standards so that it can be certified against. The United States National Institute of Standards and Technology published the voluntary AI Risk Management Framework, AI 100-1, in January 2023, organised around Govern, Map, Measure and Manage. CEN-CENELEC's JTC 21, established in June 2021, is drafting the harmonised European standards that give a presumption of conformity under the AI Act, and in October 2025 announced measures to accelerate that work. The pattern across all three is the same: governance can be specified faster than measurement can be, and the standards that would make an obligation testable are the slowest part.

A node is not a place. Drawing it on a map would assert something about the world that no stored fact supports.