Workshop
In person
LeadersGoldDiscovery

AI benchmarking and AI-powered testing

  • Date
    10 July 2026
    Timeframe
    09:00 - 12:15
    Duration
    3h 15 minutes
    • Days
      Hours
      Min
      Sec
    As AI matures, the need for standards to promote efficiency and sustainability grows more important. This workshop aims to showcase ongoing ITU-T standardization activities in the development of AI monitoring and benchmarking frameworks that support users in assessing the effectiveness, robustness, and reliability of ML/AI systems. It will also highlight emerging AI based tools that can be used to support testing procedures and automate evaluation workflows.
    Current Machine Learning (ML) and AI models are trained based on existing data and underlying asssumptions regarding their operational context. Given the dynamic nature of real-world environments, these changes can result in AI model degradation, leading to reduced predictive accuracy or diminished decision quality as data and concepts evolve. To address these issues, it is essential to identify and monitor relevant parameters using defined metrics. Accordingly, selecting appropriate monitoring strategies tailored to the use case, data properties, and organizational objectives is crucial for maintaining the long-term reliability and effectiveness oof ML/AI systems.
    The workshop focuses on the following interconnected areas:
    – AI benchmarking and monitoring to mitigate risk of AI model degradation and its performance
    – AI-enabled testing processes, including demand analysis, use case design, data generation, execution, and reporting
    – AI-driven automatic test script generation – including TTCN-based scripts and test tools
    In October 2024, the World Telecommunication Standardization Assembly (WTSA-24) adopted Resolution 101, which focuses on setting tecnical standards for artificial intelligene (AI) in telecommunications and information and communication technologies (ICTs). This workshop will explore potential new areas of study related to testing methodologies, evaluation scenarios, and bencharmking approaches for AI based models. By bringing together experts from industry, testing laboratories, the standadization community, academia, and regulatory bodies, the event will foster collaboration that can accelerate the development of new ITU-T standards for sustainable, secure, and AI-enabled digital infrastructure.
    Share this session with your network
    • Session starts in
      Days
      Hours
      Min
      Sec

    Are you sure you want to remove this speaker?