Writing the standards for artificial intelligence
Join AFNOR’s standards committee to help establish best practices for AI systems.
The core of our artificial intelligence standardization services
- A management system approach
- A study on the characterization of trust
- Future harmonized standards to support the AI Act

Why work on standardization in artificial intelligence?
Driving a car, handling a weapon, applying the right amount of fertilizer or pesticides, medical imaging diagnostics, customer service chatbots… Artificial intelligence is set to play a role in every field. Systems that rely on AI are already part of our daily lives… And this is just the beginning.
An entire industry is emerging in this sector with its many applications, making it a matter of economic sovereignty for France and Europe. In May 2021, as part of the Grand AI Challenge , the government commissioned AFNOR to assist industry stakeholders in developing voluntary standards designed to build trust in systems that use AI. Three years earlier, voluntary standardization had been identified as a key tool for this purpose . The Grand Défi IA views voluntary standardization as a tool for building trust, in support of the 2024 European AI Act, with the aim of providing designers and users of artificial intelligence systems with a dual framework—comprising both regulatory and normative standards—that offers a secure environment for developing and bringing to market their solutions and products that utilize trustworthy AI.
Be sure to check out our video "Shaping European Leadership in Artificial Intelligence: Insights from Key Organizations."

In December 2023, the first major framework standard was published: the ISO/IEC 42001 , providing guidelines for implementing a certifiable AI management system. Today, other standards are under development, including harmonized standards that provide presumption of conformity with the European Union’s Artificial Intelligence Regulation. France, through AFNOR, is working in particular on the draft European standard “Characterization of Trust for AI” within the framework of the France delegation to CEN-Cenelec JTC 21, which it co-chairs.
Terminology
Management
Risks
- ISO/IEC 23894 Guidance on AI Risk Management
- CEN-CENELEC "AI Risk Catalog"
Characteristics of Trust
- CEN-CENELEC – “Trustworthiness characterization”
- ISO/IEC standards
Data quality
- ISO/IEC 52589 series: "Data Quality for Analytics and Machine Learning"
Compliance claims
- CEN-CENELEC TR "AI Conformity Assessment"
Why Choose AFNOR Standardization
- A way to bring all your stakeholders together
- Your chance to make your voice heard
- A quest for consensus
- The pursuit of the public interest
- A step toward international standardization

