Le Poivrier Restaurant
Workshop

Inaugural meeting of ITU trustworthy AI testing and knowledge platform

In person
  • Date
    9 July 2025
    Timeframe
    15:30 - 17:30 CEST
    Duration
    2 hours
    Share this session

    The ITU Trustworthy AI Testing and Knowledge Platform (TAITKP) would work towards providing a global platform for multistakeholder collaboration and knowledge sharing at international level for research on testing, verification, and mitigation methodologies for AI systems. This initiative addresses the growing need for robust, standardized testing frameworks to evaluate AI systems across different contexts and use cases.  This is also aligned with WTSA-24 Resolution 101 on Standardization activities of the ITU Telecommunication Standardization Sector on artificial intelligence technologies in support of telecommunications/information and communication technologies.

    TAITKP aims to provide a global forum for AI experts from industry, academia, standards development bodies and Member States to disseminate latest research findings, share lessons learned and conduct capacity building on testing and evaluating AI systems. TAITKP is co-chaired by Dr Bilel Jamoussi, Deputy Director TSB, ITU and Dr Robert Trager, Oxford University and meets on quarterly basis. The main objectives are to:

    1. promote technical standards and best practices for testing, evaluation and mitigation of AI systems,
    2. Facilitate information exchange and dissemination of latest research findings and best practices on testing methodologies and verification techniques, and
    3. provide a body of knowledge for guidance and capacity building for emerging economies for evaluation and testing of AI systems.

    The inaugural meeting of the TAITKP establishes a framework for international technical collaboration on testing and verification of AI systems for alignment with trustworthy AI principles. This gathering will launch a sustainable platform for knowledge sharing, technical standards collaboration, and capacity building across nations. The meeting will focus on operationalizing the platform’s five foundational pillars:

    1. Global multi stakeholder dialogue and knowledge sharing on promoting global coordination on AI systems testing and disseminating findings of such test methods.
    2. Promote technical standards and best practices for testing and verification of AI systems
    3. Coordinate requirements for global standards for AI testing and conformity assessment
    4. Conduct studies and research on AI systems and model testing and taxonomy for Trustworthy AI Principles and test metrics
    5. Provide technical guidance and capacity building for developing countries on testing AI systems