Government of Lesotho - MICSTI
Responses
In your opinion, what outcomes would make the first Global Dialogue on AI Governance a success?
Recognition of Africa's reality of unstructured and fragmented data that is not AI-ready, according to the well-established definition of AI-ready data, which is clean, structured, and centralized in most cases. Although the last one can be easily solved with interoperability. A call for the establishment of a Framework for AI-Ready Unstructured and Fragmented Data would be highly welcomed.
From your perspective, which of the following thematic areas identified by the General Assembly Resolution 79/325 for the AI Dialogue reflect your priorities for urgent action and active engagement?
1
Open-source software, open data and open AI models;Interoperability of governance approaches;AI capacity-building;Transparency, accountability, and human oversight;
Please briefly explain your selection.
4
To deal with elements of AI development and governance, capacity building across all AI-related skills is important in order to participate meaningfully. Interoperability is key to managing decentralized databases. Additionally, human oversight should in the loop for most of the AI workflow in order to correct mistakes that can be made by AI. Lastly, in order for AI to be inclusive and accessible to everyone, AI development should be based on open standards and access across the following domains: software, data, and AI models.
In your opinion, are there any cross-cutting or emerging issues not captured by the listed themes above? If so, please explain.
AI-ready data for unstructured and fragmented datasets
How are the governance gaps and related developments/advances in the thematic areas you selected above affecting your country, region, or sector? Please highlight the most significant challenges.
On open source and open standards, Lesotho needs clear government rules requiring AI systems to use open standards for data formats, APIs, metadata, audit logs, model documentation, and portability. This calls for a regional framework on open source and open standards that we can adopt locally. Regarding interoperability, AI governance must be embedded in Digital Public Infrastructure, digital ID, eServices, data governance, and enterprise architecture. The country needs an AI-ready interoperability framework that recognises the reality of our data, which is unstructured and fragmented. Without this, AI systems may produce outcomes that are not favourable to our local conditions, as data cleaning and tagging might take a long time and also be a costly exercise to undertake. On safety and human oversight, Lesotho needs binding rules for risk classification, prohibited AI uses, mandatory impact assessments, cybersecurity review, human-rights assessment, incident reporting, and citizen redress. Human oversight must be clearly defined so officials can review, pause, override, or appeal AI-assisted decisions.
What role can the AI Dialogue play in advancing international cooperation on AI governance?
AI Dialogue that produces clear call-to-actions that regions and member states can adopt. This approach can ensure greater cooperation and interoperability among member states.
What are some of the existing initiatives, partnerships, or mechanisms that the AI Dialogue should build upon or connect with, and what added value could the AI Dialogue bring?
Clear candidates are ITU AI for Good and UN Open-Source Week.
How can different stakeholders contribute to the AI Dialogue? Please share recommendations for the format and structure of the AI Dialogue.
I think the current UN processes are adequate
Which voices, communities, or perspectives are currently underrepresented in global discussions on AI governance? How could they be included?
Youth and old people are mostly excluded.
What innovative engagement formats could most effectively foster meaningful and dynamic engagement during the AI Dialogue?
Country statements, panel discussions and ministerial round-tables
Please share examples of policies, practices, platforms, or approaches that promote effective AI governance or offer concrete solutions to addressing its challenges.
I think the Framework for AI-Ready Unstructured and Fragmented Data can help Africa accelerate its AI development and ensure AI governance frameworks address local real conditions.