On July 10, the AI for Good Global Summit hosted a seminar on standardization in the development of AI monitoring and benchmarking* systems, as well as on the related work of the ITU Telecommunication Standardization Sector (ITU‑T).

Such systems help users assess the performance, robustness, and reliability of machine learning and artificial intelligence (AI) systems.

The following interrelated areas of ITU‑T work became the central themes of the seminar:

• AI benchmarking and monitoring to mitigate risks of model degradation and performance decline

• AI‑driven testing processes — including requirements analysis, use case design, data generation, test execution, and report generation

• Automatic generation of AI‑based test scripts.

For RCC participants, this topic is particularly relevant. To confidently deploy AI in telecom, public services, and other areas, countries need common rules of the game — exactly what is being discussed on the ITU platform.

The following RCC professional community representatives shared their expertise:

• Andrey Kucheryavy, Professor, Head of the Department of Communication Networks and Data Transmission, St. Petersburg State University of Telecommunications named after Professor M.A. Bonch-Bruevich (SPbSUT)

• Anna Abramova, Ph.D. in International Economics, Head of the MGIMO AI Centre

They presented information on the Central Asian sandbox for AI benchmarking, which is being implemented jointly by the RCC, SPbSUT, and the MGIMO AI Center.

• Artem Volkov, Doctor of Technical Sciences, Associate Professor at SPbSUT, who moderated the session on AI‑based testing systems.

New tools for supporting testing procedures and automating AI‑based assessment workflows were also presented during the event.

*Benchmarking is the process of comparing indicators, methods, or results against a benchmark.

⠀

For the latest RCC news, subscribe to our official Telegram channel.