Page 114 - AI Standards for Global Impact: From Governance to Action

AI Standards for Global Impact: From Governance to Action

advanced analytics and intelligent algorithms to streamline the examination process,
addressing the challenges posed by costly, limited resources, and time-consuming
traditional methods.
ii. CAICT highlighted AI model adoption choices and preferences in various domains, and
the potential actions for the next version of the AI Readiness Report and Framework, such
as creating assessment tools, developing measurable criteria, and metrics.
iii. Global Green Growth Institute shared the framework and model its built to analyze the
co-benefits of the UN Sustainable Development Goals and provide scenario time-series
prediction visualization.
iv. Lenovo presented its ITU-supported project of AI-Powered Human-like Avatar for Sign
language in Brazil and its impacts.
v. Umgrauemeio (1.5 degree) presented its project focusing on wildland fire management.
vi. Pontificia Universidad Católica de Chile presented the progress on its Latin America AI
Index (ILIA), including the large amount of data it has collected for assessing and ranking.
Inputs from ILIA will support the development of the ITU AI Readiness toolkit.
vii. MGIMO AI Center presented AI readiness in international trade from a process-centric
point of view. Their input to the ITU AI Readiness Framework considers the importance
of finding the right place to apply AI by inputting domain-specific workflows identified by
authoritative standards in the relevant field, and quality-related risks relevant to trading AI
goods and services.
viii. DevelopMetrics emphasized the importance of preserving local data and forming a
platform with a standardized data format for future reference.
ix. The University of Applied Sciences and Arts of Southern Switzerland presented its project
on drone-supported wheelchairs with a special focus on lawful, ethical, and robust use of
AI.
x. Australia's University of Victoria shared its project using AI to provide companions for the
visually impaired population using LLMs and vision-language models. It was observed that
the evaluation of an AI application is key to assessing AI readiness.

16�3 Outcomes

A key outcome was planning the next steps toward a revised AI Readiness Report, which will
incorporate new methodologies, feedback from industry, and complement the lessons learned
from the pilot AI Readiness Plugfest with the assessment indicators. Building on these insights,
ITU will take AI readiness discussions forward by:

• Enhancing AI readiness evaluation methodologies by studying the characteristics of AI
integration in various domains.
• Exploring pathways for stakeholders to contribute to a standardized AI readiness
framework.
• Engaging the Technical Advisory Committee to provide strategic guidance and feedback.

The ITU AI Readiness Framework aims to enable countries to conduct the AI readiness self-
assessment. ITU is calling for engagement from experts in different domains to design and
refine the key factors and relevant indicators to deliver a toolkit. Learnings from plugfest project
owners are essential as they bring in regional and domain-specific points of view. Plugfest
presentations are supporting the identification of common metrics for measuring AI readiness.

102