1 of 12

Understanding

AI Policy Gaps and

Building

a Global South Network

WEBINAR 1 - Bringing together the voices from Vietnam,

Bangladesh, India, Kyrgyzstan, Ghana, South Africa and Bolivia

to build confidence in our shared context.

Date: June 25, 2025

Time: 11:00–12:30 UTC

2 of 12

Current AI Policy Landscape - VIETNAM

National Strategic Vision to 2030 (DECISION 127/ QD-TTg, 2021)

Top 4 in ASEAN, Top 50 in the world in AI; 10 global AI solution enterprises & 10 Data centers & 50 open source databases 2. Become the innovation center for AI R&D in the region. c) Contribute to creative, innovative society, effective governance, support national security, society stability and sustainable growth of economy.

The Education Infrastructure

Ground Reality/Challenges

Population: 101.3 M. 22 million students. Comp Science is required from Grade 3, since 2018
Internet penetration: broadband internet connection (4G, 5G); 78.8% population internet penetration
76.2 million social media users in (January 2025). 127 million million active cellular mobile connections (Jan. 2025)

BIG GAP between s#5 in Southeast Asia in Government AI Readiness Index (2023)school's laptop and equipments, internet connection between private vs public; big cities vs rural/mountainous areas (some schools have no computer or 2 student/1 computer a class.
Lack of trained IT teachers
No AI update in the curriculum yet
Lacking budget for equipments
Resistance of change and technology-enhanced pedagogy from teachers and parents, school leadership

#5 in Southeast Asia in Government AI Readiness Index (2023)

3 of 12

Current AI Policy Landscape - BANGLADESH

National Strategic Vision

National AI Strategy, 2023. a2i EdTech programs with UNDP. University Research (DU, BUET) - Bangla Natural Language Processing (NLP) model, MoE + UNESCO → AI Ethics.

The Education Infrastructure

Ground Reality/Challenges

DGHS eLearning Bangladesh Computer Council, a2i & Blended Education Accelerator (Muktopaath)
Gazipur Digital University plans an AI department, AI research - PhD/MS training programs.
Various digital initiatives at central and regional levels – schools, colleges, universities and medical training.

Digital infrastructure gap, Curriculum, AI literacy among faculty is missing
Low no. of AI graduates (no specific data) ~10% of all ICT graduates. Low innovation
Many digital skills courses and initiatives but private sector lacks capacity to absorb graduates.

Total Schools - 108,515 Includes all types (primary, secondary, etc.).

Public Schools - 63,041, Private/Other School - ~45,000+ (Kindergartens, NGO, English medium, madrasahs, etc)

TVET Institutes - 1,082 (About 13.2% are public, 48% are individually/family owned (private), with the rest under other categories) - These institutions are under the Bangladesh Technical Education Board (BTEB) and various ministries.

Public Universities - 55, Private Universities - 110+

Primary Education:�Grades I–V (ages 6–10), mostly in public schools.
Secondary Education:�Grades VI–XII (ages 11–17), mix of public and private schools.
TVET:�Vocational and technical education at secondary and post-secondary levels, offered by public and private institutes.
Tertiary Education:�Public and private universities offering undergraduate, graduate, and doctoral programs

The Ministry of Education in Bangladesh employs over half a million teachers across all levels, with about 384,513 in primary education alone as of 2023.
The teacher-student ratio is 1:34 as of 2023

The last major revision of the national education curriculum in Bangladesh occurred in 2021, when a competency-based curriculum was introduced for general and Madrasah education, emphasizing practical skills and real-life applications. This curriculum began phased implementation in 2022 and was scheduled to be fully in place by 2025

4 of 12

Current AI Policy Landscape - INDIA

National Strategic Vision

National Strategy for AI 2018. India AI Mission, USD 1.2 Billion budget allocation in 2024. Building National Foundational Models, 35K GPU Facility (USD 0.5 per GPU hour boosting innovation)

The Education Infrastructure

Ground Reality/Challenges

Approx 248 million students enrolled in schools and 52,538 colleges as of 2025. NEP 2020.
Universities (Public: 495, Private: 497). Question of academic integrity in AI age in schools and universities!

16.8 % (47.4 million) of school age children currently out of school 2023 - 2024. Unreliable electricity, internet and device access. Teacher Shortage. Quality of teaching.
Revised curricula (NEP) but uneven foundational digital literacy and Teacher Professional Development/missing curricula content.

Building GPU clusters doesn't address the fundamental issue that many schools and colleges lack reliable internet or basic computers.

AI mission prioritises tech infra but doesn't adequately address curriculum, massive teacher training and capacity building required.

�As of the 2023-24 academic year, India employs 98 lakh (9.8 million) teachers in its school education system, covering 14.72 lakh schools and serving 24.8 crore (approx 250 million) students, approximately 16 lakh (1.6 million) teachers in higher education institutions (HEIs), including universities, colleges, and stand-alone institutions

The IndiaAI Mission represents a "top-down" technological approach. It does not align with the "bottom-up" needs of India's education system (and hence the relevance of this Network!

“Fundamental Disconnect” - Minimal direct impact on K-12 (Kindergarten - class 12) education due to infrastructure gaps, might be potential benefits for higher education institutions, private universities and research, but it does risk of creating further stratification or algorithmic divide (similar to digital divide) between well-resourced and under-resourced education institutions and students.��With this pace of development and an overly ambitious target of 10 months (Jan, 2025) to a foundational LLM India might see a possible trickle-down effects on the education sector if foundational models become more accessible to public in the next 3 to 7 years. there is hence a potential for innovative educational applications, but primarily in urban areas. �It would hence be safe to say that the challenges of access and inclusion in rural and economically disadvantaged areas will persist in the path to the digital transformation of India.

On a Long Term (5 -10 years down the line) through collective initiatives like this network (catalyst function) AI could have a positive impact on education if accompanied by policy framework for comprehensive infrastructure development, edtech ecosystem boost, parallel investments in curriculum and teacher training.

5 of 12

Current AI Policy Landscape - KYRGYZSTAN

National Strategic Vision

Digital Kyrgyzstan 2024-2028”, “National Development Program of Kyrgyzstan until 2030” and government announcements via the National Council on AI (Feb 2025). Creation of specialized STEM schools and lyceums.

The Education Infrastructure

As of 2023, around 4,989 primary and secondary schools. 58 universities. 25000 teachers
TVET is a part of the upper secondary and post-secondary non-tertiary education

Ground Reality/Challenges

Internet connectivity in schools, language diversity, physical infrastructure, teacher professionalism and motivation, students’ learning outcomes, and STEM curriculum, parent involvement.

National Council for AI Development (as part of the Digital Transformation Concept for 2024–2028). The National Council’s tasks include:

● Developing projects in the field of AI.

● Enhancing the country’s scientific and educational potential.

● Strengthening Kyrgyzstan’s international position in the AI sphere

National Development Program of Kyrgyzstan until 2030, Section 3, New Horizons^[1]

Development of Human Capital - Key Levers - Development of AI competencies and STEM education:

- Reform of educational programs by introducing the basics of programming, robotics and artificial intelligence into the school curriculum

- Development and implementation of specialized courses in AI, Big Data and machine learning in universities and professional colleges

- Creation of specialized STEM schools and lyceums with in-depth study of technical sciences and engineering disciplines

- Organization of competitions and olympiads in AI and robotics to identify talented students and specialists

- Formation of a balanced system that promotes the development of intellectual property, innovation and creativity, as well as stimulation of inventive and creative activities

^[1] National Development Program of Kyrgyzstan until 2030

6 of 12

Current AI Policy Landscape - GHANA

National Strategic Vision

Republic of Ghana National Artificial Intelligence Strategy: 2023-2033 emphasises under Pillar 1: Expand AI Education and Pillar 2: Empowering Youth & Jobs (Promote training courses for teachers

The Education Infrastructure

Ghana’s education system is structured as a 6-3-4-4 cycle. ~15,000 schools, ~200,000 Teachers. 97.9% Primary enrollment
ICT in Education developed in 2003, reviewed 2015, third review in progress

Ground Reality/Challenges

National Policy gives no specifics on pedagogical methods, localization, or incentives for in-service teachers. National EdTech Strategy under development
No mention of aligning youth training programs with local tech industry needs. Math and english learning outcome remain low. Teacher Training, Edtech and Connectivity disconnect.

7 of 12

Current AI Policy Landscape - SOUTH AFRICA

National Strategic Vision

AI Policy Framework, Oct 2024. Section 6.2: Develop supercomputing infrastructure and invest in 5G and fiber networks. 6.1 mentions “incorporate AI into educational curricula” and “specialised training programs”

The Education Infrastructure

Foundation (Grades R–3), Intermediate (4–6), Senior (7–9), and Further Education (10–12). 25,000 schools and 26 Universities. 400,000+ teachers
Curriculum and Assessment Policy Statement (CAPS), 2012 - Revision in 2022.

Ground Reality/Challenges

Schools face shortages of classrooms, learning materials, language diversity (12 national languages) and basic amenities
Between 50-70% schools lack reliable internet access. Deep urban-rural divide.

Education access has improved, however the quality of teaching, infrastructure, and digital literacy, especially in rural areas remain a big challenge. These issues are directly relevant to the ambitions of the National AI Policy Framework.

AI Policy Sections (Specific to Education) and its Connection to the Current Education Landscape

6.1 Talent/Capacity Development

The uneven quality of foundational education, digital literacy, and teacher training (in-service and new teacher) means that South Africa must urgently invest in upskilling teachers, updating curricula to include AI and digital skills, and ensuring all learners have access to quality STEM education. Without addressing these gaps, the country risks a shortage of AI-ready talent.

6.2 Digital Infrastructure

The digital divide, especially between rural and urban schools, directly impacts the ability to roll out AI education and leverage digital tools. Upgrading connectivity, devices, and digital learning platforms in all schools is essential to create an enabling environment for AI innovation.

6.3 Research, Development, and Innovation

A strong foundation in STEM, digital literacy, and higher education is necessary to support AI research and innovation. Partnerships between schools, universities, and industry as well as targeted support for research centres will be critical, but these depend on addressing the current quality and infrastructure gaps in the basic education system.

POLICY - AI Policy Framework, Oct 2024. Section 6.2: Develop supercomputing infrastructure and invest in 5G and fiber networks. 6.1 mentions “incorporate AI into educational curricula” and “specialised training programs”��EDUCATION - Foundation (Grades R–3), Intermediate (4–6), Senior (7–9), and Further Education (10–12). 25,000 schools and 26 Universities. 400,000+ teachers

Curriculum and Assessment Policy Statement (CAPS), 2012 - Revision in 2022.

�REALITY - Schools face shortages of classrooms, learning materials, language diversity (12 national languages) and basic amenities

Between 50-70% schools lack reliable internet access. Deep urban-rural divide.

8 of 12

Deep national crisis: Political, economical, and social crisis override long-term planning.
Mindset barriers: Strong resistance to change and low incentives to adopt/trust tech.
Language exclusion: Tools and content often ignore Bolivia’s multilingual reality.
Low English: Limits access to state-of-the-art research, training, and resources.
Costly, poor internet: Bolivia, one of the worst speed-to-cost internet ratios in LATAM.

Centralized control: Ministry of Education defines curriculum and policy; municipalities only implement.
Infrastructure gaps: Many schools lack electricity, internet, or devices—especially in rural areas.
Curriculum reforms highlight interculturality, but AI and digital content are absent.

Current AI Policy Landscape - BOLIVIA

National Strategic Vision

Bolivia lacks a national AI strategy or education-focused digital policy. Efforts remain fragmented, with limited public visibility or coordination.
Private and multilateral actors are filling the vacuum, but without systemic integration.

The Education Infrastructure

Ground Reality/Challenges

Quantitative Data (Recent Years)

Enrollment:

Net enrollment in primary education is above 85%.
Secondary completion and transition rates are lower, especially in rural areas and among indigenous populations.

Teachers:

The Ministry of Education maintains a national teacher registry and pays teachers directly, but distribution and quality vary by region

Universities:

10 public and 23 private universities, with the University of San Andrés (La Paz) and San Francisco Xavier (Sucre) being the largest and oldest, respectively.

TVET

Technical/vocational education is available at the secondary level and through higher vocational institutes, but is largely urban and fragmented, with limited coordination and access for rural youth

The government and international partners (including the World Bank) have promoted digital learning platforms, teacher ICT training, and the integration of technology into classrooms, but challenges remain in scaling and sustaining these efforts nationwide.

Policy Focus: Current policy and donor attention is on foundational digital skills, infrastructure, and teacher digital literacy, rather than advanced AI competencies.
Potential for Future Integration: As digital transformation deepens and regional AI activity grows, Bolivia may move toward more structured AI competency integration in education.
Digital Divide: Significant urban-rural and socioeconomic divides in access to digital infrastructure limit the potential for widespread AI education.