AI Trust Alliance

Base specification

IEEE, Positive AI, IRT SystemX and VDE have jointly developed a specification to serve as a basis for assessing the trustworthiness of AI systems. The specification translates high-level trustworthy principles into concrete indicators that address both technical and organizational practices. It can be used as a first step toward certification or as a self-assessment tool to help organizations understand where they and/or their products stand in relation to Trustworthy AI practices. 

The joint specification is based on 7 core principles which underpin the EU AI Act’s approach to Trustworthy AI:

  • Accountability  
  • Human Agency & Oversight 
  • Technical Robustness & Safety 
  • Privacy & Data Governance 
  • Transparency 
  • Diversity, Non-Discrimination & Fairness 
  • Societal & Environmental Well-Being 

Version 1.0 was published in March 2025.

  • The specification defines a method to assess the trustworthiness of AI products and services. More details and download ->
  • Version 1.0 was used in the development of Germany’s national MISSION KI Quality Standard.

Version 2.0 draft was published in December 2025. It introduces several key updates: 

  • A new core principle of Accountability has been added, including sub-principles such as Auditability and Impact Assessment & Mitigation Measures, and the Human Agency & Oversight principle from Version 1.0 updated accordingly. 
  • Indicators for AI system assessment have been restructured into two categories, to cover specificities in the AI system lifecycle: AI developed systems and AI systems in operation. 
  • Most importantly, Version 2.0 introduces the concept of “evidence” to support the measurement of indicators, making the evaluation more actionable and verifiable. 

Interested?

Access the joint “Specification for the Assessment of the Trustworthiness of AI Systems” working draft Version 2.0, including the updated principles, indicators, and evidence framework. More details and download ->

We welcome your feedback on this working draft.  

Please share your comments and suggestions through this form ->