Correspondence Group on SG13 future direction (CG-SG13ftr) was set up by SG13 on 3 March 2025 (SG13 meeting March 2025, Geneva). The Terms of Referece is available in TD23-R1/PLEN.
- Terms of reference: Under the auspices of ITU-T SG13, the CG-SG13ftr will carry out the preliminary analysis and consideration to support that SG13 Questions effectively and efficiently conduct standardization works on future networks. The objectives are to discuss the handling of study topics in each Question with the change of organizational structure of SG13 in Study Period 2025-2028.These activities include:
‒ Analysis of the current Questions’ activities (e.g., study topics, average numbers of received contributions);
‒ Survey on new topics on future networking (e.g., study items relevant to other SDOs activities
‒ Exchange opinions towards making SG13 activities more optimal
- Co-conveners: Mr Kazunori Tanikawa (NICT, Japan) and Ms Rim Belhassine-Cherif (Tunisie Telecom, Tunisia)
- Mailing list: t25sg13cgftr@lists.itu.int
- Mailing list: Archive
- Lifetime: end of 2026
Correspondence Group for datasets applicable for AI/ML in networks (CG-datasets for AI/ML in networks) was set up by SG13 on 15 July 2022 (SG13 meeting July 2022, Geneva). Since then its lifetime was extended twice (reference documents TD51/PLEN , TD265-R1/PLEN and TD93/PLEN). Updated (currently in force) Terms of Reference may be found in TD93/PLEN. -
Terms of reference: Under the auspices of ITU-T SG13, preliminary activities will be carried out to provide guidelines on a standardization approach for datasets applicable for AI/ML in networks, including
a) Study standardization requirements of datasets applicable for AI/ML in networks and applications and services enabled by networks;
b) Conduct a gap analysis of related standardization efforts;
c) Extend the analysis on Large AI models by considering also action models and other foundational models.
d) Identify directions for, and recommend, a standardization approach for datasets applicable for AI/ML in networks and applications and services enabled by networks (including allied aspects such as data models, metadata and impact on the networks and applications and services enabled by networks of AI/ML techniques in relation with this study) with associated potential work items;
e) As focus of the CG-datasets activities in this extended lifetime, organize on a periodic basis, possibly thematic, webinars/workshops. Examples of themes include, but are not limited, data quality, datasets for AI/ML applications, use cases, models, toolsets, open source and synthetic datasets. The targeted periodicity of these webinars/workshops will be 4 to 6 months. The workshops’ organization and holding, in terms of workshop themes and invited speakers, is expected to have the support of ITU-T SG13 and other interested ITU-T SGs and SDOs, as well as other experts in the AI/ML domain.
f) If judged appropriate, the development of new CG-datasets deliverables based on the above-mentioned focus of the CG-datasets activities in this extended lifetime is not excluded, as well as the enhancement of the CG-datasets deliverables submitted to the SG13 meeting, Tashkent, 28 Oct – 6 Nov 2025 (i.e., the CG-datasets Technical Report - Second Edition, and the AI/ML datasets inventory - First Edition).
-
Co-conveners: Mr Marco Carugi (marco.carugi@gmail.com), Antonio De Domenico (antonio.de.domenico@huawei.com) and Mr Vishnu Ram Ov (vishnu.n@ieee.org)
- Mailing list: t25sg13cgdatasets@lists.itu.int
- Mailing list: Archives
- Documents
- Outputs (July 2024) - technical report ; (October 2025) - technical report; AI/ML datasets inventory - First edition
- Lifetime: October 2026