Executive Summary
Artificial
intelligence has moved from experimental technology to a general-purpose driver
of economic transformation, and development economics is being rewritten as a
result. Global corporate AI investment reached $252.3 billion in 2024 and, per
Stanford's 2026 AI Index, surged to $581.7 billion in 2025 — yet more than
three-quarters of that private capital remains concentrated in the United
States a
lone. This asymmetry sits atop deeper divides: 2.2 billion people
remain offline globally, five-sixths of them in low- and middle-income
countries, and only 23% of people in low-income countries use the internet
compared with 94% in high-income economies. At the same time, IMF and ILO
research shows AI's labour-market exposure is lower in poorer countries (around
26–28% versus 60% in advanced economies) — a "double-edged" finding,
since lower exposure also signals weaker readiness to capture AI's productivity
dividend. This article examines how development economics is evolving in
response: from digital public infrastructure in India and Estonia to AI-enabled
health logistics in Rwanda, precision agriculture, and algorithmic public administration.
It argues that the decisive variable is not technological access but
institutional capability — governance, human capital, data infrastructure, and
regulatory maturity. Without deliberate policy design, AI risks entrenching a
new hierarchy of nations; with it, AI can become one of the most powerful
accelerants of inclusive growth since the Green Revolution. The article closes
with concrete recommendations for governments, development banks, the private
sector, and universities.
Introduction
Development
economics has reinvented itself roughly once a generation — from
capital-accumulation models in the 1950s, to structural transformation and
human capital theory, to institutional economics and the Sustainable Development
Goals. Each reinvention responded to a technology or a crisis that changed what
growth required. Artificial intelligence is the newest such force, and arguably
the most consequential since electrification and telecommunications reshaped
industrial economies a century ago.
The scale of
the shift is now measurable. <cite index="8-1">US private AI
investment reached $285.9 billion in 2025, more than 23 times the $12.4 billion
invested in China</cite>, according to Stanford's 2026 AI Index — a
concentration of capital that has few historical parallels in general-purpose
technology diffusion. Corporate use of AI has become mainstream: organisations
reporting AI adoption in at least one business function jumped from 55% to 78%
between 2023 and 2024. PwC's widely cited "Sizing the Prize" study
projects that AI could add $15.7 trillion to global GDP by 2030 — more than the
current combined output of China and India — split between productivity gains
and consumption effects.
For developing
economies, this moment carries a genuine duality. On one hand, AI offers a
plausible leapfrogging pathway: countries that never built extensive fixed-line
telephony went straight to mobile; nations without dense networks of specialist
doctors are now piloting AI diagnostics; tax authorities without large audit
bureaucracies are using machine learning to detect evasion. On the other hand,
the same digital and institutional gaps that limited earlier waves of
technology diffusion — connectivity, electricity, skills, data governance
capacity — now determine who can absorb AI's benefits at all. The ILO and World
Bank's most recent joint analysis found that <cite
index="27-1">exposure to generative AI is higher in advanced
economies, particularly in clerical and professional occupations, while
developing countries, though less exposed overall, face structural constraints
that limit their ability to benefit from the technology</cite>. Lower
exposure, in other words, is not automatically good news — it can equally
reflect exclusion from AI's upside.
This article
treats AI as what economists call a general-purpose technology: one whose value
depends less on the technology itself than on the complementary investments —
infrastructure, skills, institutions, and regulation — that societies build
around it.
The Evolution of Development Economics
Classical
development economics centred on capital accumulation and industrialisation
(Lewis, Rostow, Solow). By the 1990s, the field had absorbed human capital
theory (Schultz, Becker), recognising that education and health were productive
investments, not merely welfare spending. The 2000s brought an institutional
turn — Acemoglu, Robinson, North, and others demonstrated that property rights,
rule of law, and governance quality explain cross-country income divergence
better than factor endowments alone. The Millennium Development Goals and their
successor, the Sustainable Development Goals, then reframed development as a
multidimensional project spanning poverty, health, education, gender equality,
and environmental sustainability.
Digital
transformation added a further layer from the 2010s onward: mobile banking,
e-government, and broadband access became development variables in their own
right, tracked by the World Bank's Digital Adoption Index and the ITU's
connectivity indicators. AI represents the next stage in this lineage — not a
replacement for earlier theory, but an amplifier that can accelerate or distort
structural transformation, human capital returns, and institutional performance
depending on how it is deployed. Where earlier technologies primarily augmented
physical labour, AI increasingly augments — and in some tasks substitutes for —
cognitive and administrative labour, which changes the calculus for
service-based development strategies that many middle-income countries have
relied upon.
AI as a New Driver of Economic Development
AI's growth
contribution operates through several transmission channels relevant to
developing economies:
Productivity
and firm performance. Machine
learning and generative AI tools lower the cost of tasks such as translation,
document processing, credit assessment, and quality control — functions
historically constrained by scarce skilled labour in low-income settings. OECD
enterprise survey work consistently finds that firms combining digital adoption
with complementary investments in worker training and organisational change
capture substantially larger productivity gains than firms adopting technology
alone.
Agriculture. Satellite-based crop monitoring,
AI-driven pest and disease detection, and precision irrigation are being
piloted across South Asia and Sub-Saharan Africa, where agriculture still
employs a large share of the workforce. These tools do not require farmers to
own advanced hardware; many operate through SMS or basic smartphone interfaces
layered on public satellite data.
Healthcare. AI-assisted diagnostics are extending
specialist-equivalent screening — for tuberculosis, diabetic retinopathy, and
maternal health risk — into areas with acute shortages of radiologists and
specialists. Combined with logistics innovations such as Rwanda's drone-based
blood delivery network, discussed below, AI is reshaping the economics of
health service delivery in low-density, low-infrastructure settings.
Financial
inclusion. AI-enhanced
credit scoring uses mobile-money transaction histories and alternative data to
extend credit to populations without formal credit histories — a channel
already validated at scale by Kenya's mobile-money ecosystem and now being
layered with machine-learning underwriting across East Africa and South Asia.
Public
administration. Tax
authorities, customs agencies, and social-protection programmes are adopting
machine learning for fraud detection, beneficiary targeting, and revenue
forecasting, potentially improving state capacity without proportional
increases in administrative headcount — a historically binding constraint in
many developing states.
Education and
labour markets are treated in
dedicated sections below, given their centrality to how AI's gains are — or are
not — broadly shared.
Opportunities for Developing Countries
Agriculture. Precision agriculture platforms
combining satellite imagery, soil sensors, and machine learning are being
deployed to optimise input use and forecast yields, with climate-smart
applications increasingly integrated into national adaptation strategies as extreme
weather variability rises. The World Bank and CGIAR research centres have
documented meaningful yield and input-efficiency gains from digital advisory
services in smallholder contexts, though scaling beyond pilot programmes
remains uneven.
Healthcare. Beyond diagnostics, AI-enabled
logistics is already operating at national scale in parts of Africa. In Rwanda,
<cite index="55-1">Zipline's drone network was established as
the national blood-delivery service in a country where more than 80% of the
population lives in rural areas</cite>, and independent (though not yet
peer-reviewed) research from Wharton researchers found the programme was
associated with a substantial reduction in maternal deaths from postpartum
haemorrhage in the areas it serves. The programme illustrates how AI-optimised
logistics can compensate for weak physical infrastructure rather than waiting
for that infrastructure to be built first.
Education. Adaptive learning platforms show
promise for addressing teacher shortages and heterogeneous classroom skill
levels, particularly for foundational literacy and numeracy — UNESCO's global
education monitoring work has flagged both the potential and the risk that
unequal device and connectivity access could widen rather than close learning
gaps if deployed without complementary investment in teacher training and
offline-capable tools.
Public
administration. India's
digital public infrastructure stack offers the most extensively documented case
globally. <cite index="42-1">The World Bank estimates that
Aadhaar and related reforms save the Indian government over $1 billion per year
in leakage</cite>, while independent econometric analysis estimates
<cite index="41-1">UPI's contribution at roughly 3.4% of
India's annual GDP and Aadhaar's contribution in the range of 2.5% to 4.3% of
GDP</cite>, driven respectively by reduced transaction costs and more
efficient direct benefit transfers. This "digital public
infrastructure" model — interoperable identity, payments, and
data-exchange layers — is now being referenced by the G20 and replicated, in
adapted form, across parts of Africa and Southeast Asia.
SMEs. AI-powered marketing, inventory
optimisation, and export-readiness tools are lowering the fixed costs that
historically excluded small firms from formal digital markets. OECD and World
Bank enterprise survey data consistently show a positive correlation between
digital-tool adoption and labour productivity among SMEs, though causality runs
in both directions — more productive firms also adopt technology faster — which
complicates simple policy conclusions.
Financial
inclusion. Kenya's
M-Pesa, which <cite index="48-1">now serves over 40 million
users and is expanding regionally</cite>, remains the reference case for
mobile-money-enabled inclusion, and AI-based alternative credit scoring built
on top of such transaction data is increasingly extending formal credit to
previously unbanked populations across East Africa and South Asia.
Challenges and Risks
The opportunities
above are real, but so are the constraints that could turn AI into a new axis
of global inequality rather than a bridge across the old one.
The digital
divide remains the binding constraint. <cite index="39-1">Roughly 2.2 billion
people remain offline globally, most of them in low- and middle-income
countries</cite>, and <cite index="38-1">only 23% of
people in low-income countries use the internet compared with 94% in
high-income countries</cite>. Meaningful connectivity — not just coverage
— is a further gap: <cite index="40-1">only 4% of people in
low-income countries have 5G coverage compared with 84% in high-income
countries, and a typical user in a wealthy country generates nearly eight times
more mobile data than a user in a low-income country</cite>. AI cannot
compensate for infrastructure that does not exist.
Job
displacement and the "white-collar bypass" risk. IMF analysis finds AI exposure of
roughly 60% of jobs in advanced economies, 40% in emerging markets, and 26–28%
in low-income countries. Superficially reassuring for poorer countries, this
masks a specific vulnerability: recent ILO–World Bank research on 135 countries
found that <cite index="27-2">workers in jobs vulnerable to
automation are often already online even in low-income settings, meaning job
losses could occur relatively quickly, and these are frequently the
higher-quality clerical and administrative jobs that have historically offered
a pathway to decent work</cite> — particularly for women and young
entrants to the labour market. Meanwhile, workers in roles with the greatest
potential productivity gains from AI frequently lack the connectivity needed to
benefit.
Algorithmic
bias and data governance. Machine-learning models trained predominantly on data from high-income,
English-language contexts risk performing poorly — or unfairly — when applied
to under-represented populations, languages, and use cases, a concern raised
repeatedly in OECD and UNESCO responsible-AI guidance. Many developing
countries also lack the data-protection and algorithmic-accountability
frameworks now standard in the EU and OECD members, creating regulatory gaps
that can expose citizens to poorly governed AI deployment.
Infrastructure
and energy demand. AI's
computational intensity carries a growing energy and water footprint; expanding
compute capacity in regions with unreliable electricity grids — a majority of
low-income countries — raises both cost and environmental questions that are
only beginning to be systematically studied by the IEA and Stanford's AI Index.
Cybersecurity
and institutional capacity. Weaker regulatory and cybersecurity infrastructure in many developing
states increases exposure to AI-enabled fraud, disinformation, and
critical-infrastructure risk, at precisely the moment state capacity is most
needed to manage AI's labour-market and fiscal consequences.
International Case Studies
Singapore has pursued a whole-of-government AI
strategy under its Smart Nation programme since 2014, integrating AI into
public housing management, healthcare, and logistics, underpinned by
consistently ranking among the top nations globally on the UN E-Government
Development Index. The lesson for developing economies is the value of
sustained, coordinated political commitment over a decade or more, rather than
a single flagship initiative.
South Korea has combined an early national AI
strategy with deep public investment in semiconductor manufacturing,
positioning the country as both a producer and consumer of AI infrastructure.
Its consistent top-five ranking on the UN E-Government Development Index
reflects decades of investment in digital public administration that predates
its AI strategy — underscoring that AI strategies succeed on top of, not instead
of, foundational digital government capacity.
Estonia offers perhaps the most replicable
governance model for smaller and lower-capacity states. With a population of
just 1.3 million, <cite index="73-1">Estonia became the first
country to declare its government services fully digitalised in
2024</cite>, built on the X-Road decentralised data-exchange platform and
mandatory digital identity. <cite index="71-1">Digital
signatures alone are estimated to save around 2% of Estonia's GDP
annually</cite>. The lesson is architectural: interoperable "digital
rails" (identity, data exchange, signatures) built once can support
successive waves of e-government and, later, AI-enabled services — a template
increasingly referenced by the World Bank for smaller developing states.
India demonstrates that digital public
infrastructure can scale to over a billion people. Its Aadhaar–UPI stack,
detailed in Section 3, has become a reference model actively being exported
through South-South cooperation initiatives, including partnerships with
African governments on digital identity and payments.
Rwanda shows how a low-income country can
leapfrog physical infrastructure gaps using AI-optimised logistics rather than
waiting to build conventional road and cold-chain networks — its Zipline drone
network for medical deliveries is the most extensively documented example
globally of AI-enabled healthcare logistics at national scale.
Kenya built the foundational mobile-money
infrastructure — M-Pesa — that much of the current generation of AI-enhanced
fintech and credit-scoring innovation across Africa now builds upon,
illustrating how earlier-generation digital infrastructure creates the data
layer that AI applications subsequently exploit.
Vietnam,
Brazil, China, and the United Arab Emirates each illustrate distinct paths: Vietnam's
manufacturing-linked digital upgrading strategy, Brazil's use of digital
government platforms for large-scale social protection targeting, China's
state-directed AI industrial policy combined — per Stanford's Index — with
government guidance funds estimated to have deployed roughly $184 billion into
AI firms outside conventional private-investment channels, and the UAE's early
creation of a dedicated ministerial AI portfolio aimed at economic
diversification beyond hydrocarbons. Together, these cases confirm there is no
single template — but common threads include sustained political commitment,
interoperable digital infrastructure built before AI-specific investment, and
explicit attention to inclusion from the outset rather than as an afterthought.
AI and the Future Labour Market
Occupational
categories most exposed to AI globally are concentrated in clerical work, data
processing, basic customer service, and routine professional tasks such as
first-pass legal or financial document review — precisely the mid-skill service
jobs that many developing and emerging economies had targeted as engines of
structural transformation (business process outsourcing, call centres,
back-office services). This is a meaningful strategic risk for countries whose
growth models leaned on labour-cost arbitrage in exactly these functions.
Jobs likely to
expand include those requiring physical presence combined with judgment
(skilled trades, healthcare delivery, elder and childcare), AI-complementary
technical roles (data annotation, AI system oversight, model auditing), and
functions requiring contextual, relational, or creative judgment that current
AI systems handle poorly.
IMF analysis is
consistent in finding that college-educated workers move more easily from
AI-exposed to AI-complementary roles than workers with less formal education,
and that older workers face greater adjustment difficulty regardless of income
level — implying that skills policy, not just connectivity policy, will
determine whether AI narrows or widens inequality within countries, not only
between them.
For
universities and technical and vocational education and training (TVET) systems
in developing countries, the implication is a shift from teaching static
technical skills toward teaching adaptability: data literacy, critical
evaluation of AI outputs, and hybrid human-AI workflows, alongside continued
investment in the foundational literacy and numeracy that remain prerequisites
for any higher-order skill. Lifelong learning infrastructure — currently
underdeveloped in most low- and middle-income countries — is likely to become
as important a piece of social protection architecture as pensions or unemployment
insurance.
Implications for Development Economists
The
profession's own tools are changing alongside the economies it studies.
Satellite imagery and machine learning now allow researchers to estimate poverty,
agricultural output, and even conflict risk in areas where household survey
data is sparse or outdated — a capability increasingly used by the World Bank
and academic researchers to fill data gaps in low-capacity statistical systems.
Big data from mobile-phone networks and digital payment platforms is being used
for high-frequency economic monitoring, offering a level of real-time
granularity that traditional annual surveys cannot match.
Machine
learning is also reshaping impact evaluation: predictive analytics can improve
the targeting of randomised controlled trials, while causal machine-learning
methods are being adopted to detect heterogeneous treatment effects across
subpopulations — refining, rather than replacing, the evidence-based policymaking
tradition established by the "credibility revolution" in development
economics over the past two decades. Behavioural economics, too, is being
operationalised at scale through AI-personalised nudges in digital government
and financial-inclusion platforms, raising new questions about consent and
manipulation that the profession is only beginning to grapple with.
For development
economists, digital and AI fluency is becoming a core professional competency
alongside econometrics and field research design — not a specialised sideline.
Policy Recommendations
Governments
(short-term): Prioritise
foundational connectivity and electricity infrastructure before AI-specific
investment; establish baseline data-protection and algorithmic-accountability
legislation; pilot AI in tax administration and social-protection targeting,
where returns on state capacity are highest and risks are most containable.
Governments
(long-term): Build
interoperable digital public infrastructure (identity, payments, data exchange)
as a platform for successive AI applications, following the Estonia and India
models; integrate AI literacy into national curricula; establish independent AI
oversight bodies with real enforcement capacity.
Development
banks (World Bank, IMF, ADB, IFC): Expand concessional financing explicitly tied to digital
and AI-readiness infrastructure in low-income countries; fund rigorous impact
evaluation of AI pilots before large-scale replication; support regional
data-infrastructure pooling for countries too small to build AI capacity alone.
UN agencies
(ILO, UNDP, UNESCO, ITU): Lead global standard-setting on responsible AI use in labour markets and
education; monitor displacement in AI-exposed, developing-country service sectors
(BPO, call centres) and support transition programmes proactively rather than
reactively.
Private sector
and technology companies: Invest in low-bandwidth and offline-capable AI tools suited to
infrastructure-constrained settings; support local-language model development,
given the concentration of frontier AI training data in a small number of
high-resource languages; partner with governments on data-protection-compliant
deployment rather than extractive data models.
Universities
and researchers: Build local AI
research capacity rather than relying solely on imported models; integrate AI
ethics and governance into economics and public policy curricula; prioritise
applied research on AI's distributional effects within developing economies,
where evidence remains thin relative to advanced-economy research.
SMEs: Adopt AI incrementally, starting with
low-cost, high-return functions such as customer communication and inventory
management, supported by public digital-adoption grant programmes where available.
International
donors: Shift from
siloed digital-development projects toward supporting the foundational
"digital rails" — identity, payments, connectivity — that make later
AI investment productive, avoiding the trap of funding flagship AI pilots that sit
atop inadequate infrastructure.
The Future Outlook
Several
emerging themes will shape the next decade of development economics. Progress
toward increasingly capable and general AI systems raises open questions about
how quickly labour-market disruption could accelerate beyond the gradual pace
assumed in most current IMF and ILO scenarios — a genuine area of uncertainty
that this article does not claim to resolve. AI applications targeted directly
at the Sustainable Development Goals — climate adaptation modelling, disease
surveillance, early-warning systems for extreme weather — represent some of the
highest-value, lowest-controversy use cases for development-focused AI
investment. Digital public infrastructure is likely to keep expanding as the
preferred architecture for developing-country technology strategy, given its
demonstrated returns in India and Estonia. Smart-city and precision-agriculture
applications will continue to scale, particularly where they can be layered
onto existing mobile and satellite infrastructure rather than requiring new
fixed infrastructure. Concepts such as "green AI" (minimising the
energy footprint of AI systems) and "human-centred AI" (designing
systems around human oversight and complementarity rather than pure automation)
are moving from academic discussion into procurement standards at institutions
including the OECD and World Bank. Finally, global governance of AI — spanning
data flows, model accountability, and cross-border regulatory coordination —
remains fragmented, and developing countries currently have limited voice in
the international bodies shaping these rules, a governance gap with direct
economic consequences.
Conclusion
The future of
development economics will not be determined by the sophistication of AI models
alone. It will be determined by whether institutions are capable enough, human
capital deep enough, and governance frameworks fair enough to convert AI's raw
technical potential into broadly shared prosperity. History offers a caution
here: earlier general-purpose technologies — the steam engine, electrification,
the internet — each generated enormous aggregate wealth while also
concentrating early gains among those already best positioned to capture them.
AI is on a similar trajectory, and the data reviewed in this article — from
investment concentration to connectivity gaps to differential labour-market
exposure — suggest the divergence risk is real, not hypothetical.
But the same
evidence also points to a genuine and achievable alternative path. Rwanda did
not need decades to build hospital-grade cold-chain infrastructure before
AI-optimised drone logistics began saving mothers' lives. India did not need to
wait for universal banking penetration before Aadhaar and UPI extended
financial access to hundreds of millions of people. Estonia, with 1.3 million
citizens, built government infrastructure now studied by nations a hundred
times its size. These are not accidents; they are the product of deliberate
institutional design applied early and consistently.
The call to
action is straightforward, if not easy: governments must treat digital and
AI-readiness infrastructure as a development priority on par with roads and
electricity; development banks and donors must fund the unglamorous
foundational layers rather than only the flagship pilots; universities must
build local capacity rather than permanent dependence on imported models; and
the private sector must be a genuine partner in inclusive deployment, not
merely a vendor. AI will not, by itself, close the gap between rich and poor
nations. Applied with the same rigour, evidence discipline, and institutional
patience that has defined development economics at its best, it can help close
that gap faster than any technology before it.
Comparison
Table — Traditional vs. AI-Driven Development Economics
|
Dimension |
Traditional Development Economics |
AI-Driven Development Economics |
|
Data sources |
Household
surveys, national accounts, periodic censuses |
Real-time
satellite, mobile, and transaction data supplementing surveys |
|
Decision-making |
Rule-based
policy design, expert judgment |
Predictive
analytics and machine-learning-informed policy design |
|
Policy design |
Sector-specific,
often siloed programmes |
Cross-sectoral,
platform-based digital public infrastructure |
|
Productivity
drivers |
Capital
accumulation, labour, physical infrastructure |
Data,
algorithms, human-AI complementarity, digital infrastructure |
|
Evaluation
methods |
Randomised
controlled trials, difference-in-differences |
RCTs
augmented with causal machine learning and high-frequency data |
|
Citizen
engagement |
In-person
service delivery, periodic consultation |
Digital
platforms, personalised (and potentially algorithmically mediated) engagement |
|
Innovation
capacity |
Concentrated
in formal R&D institutions |
Increasingly
distributed via open-source models and low-cost digital tools |
References (APA 7th Edition)
Acemoglu, D.,
& Restrepo, P. (2022). Tasks, automation, and the rise in US wage
inequality. Econometrica, 90(5), 1973–2016.
Asian
Development Bank. (2021). Digitalization and economic performance of two
fast-growing Asian economies: India and the People's Republic of China.
ADB.
Cazzaniga, M.,
Jaumotte, F., Li, L., Melina, G., Panton, A. J., Pizzinelli, C., Rockall, E.,
& Tavares, M. M. (2024). Gen-AI: Artificial intelligence and the future
of work (IMF Staff Discussion Note SDN/2024/001). International Monetary
Fund.
https://www.imf.org/-/media/files/publications/sdn/2024/english/sdnea2024001.pdf
Centre for
Public Impact. (2024). e-Estonia, the information society since 1997.
https://centreforpublicimpact.org/public-impact-fundamentals/e-estonia-the-information-society-since-1997/
Chui, M.,
Manyika, J., & Miremadi, M. (2023). Generative AI and the future of work
in America. McKinsey Global Institute.
CSEP (Centre
for Social and Economic Progress). (2025). Measuring the economic impact of
India's digital public infrastructure: An assessment.
https://csep.org/discussion-note/measuring-the-economic-impact-of-indias-digital-public-infrastructure-an-assessment/
Gmyrek, P.,
Berg, J., & Bescond, D. (2023). Generative AI and jobs: A global
analysis of potential effects on job quantity and quality (ILO Working
Paper 96). International Labour Organization.
Gmyrek, P., et
al. (2025). Generative AI and jobs: A refined global index of occupational
exposure (ILO Working Paper 140). International Labour Organization.
International
Labour Organization. (2026). Work transformed: Promise and peril of AI.
ILO.
https://www.ilo.org/sites/default/files/2025-07/ilo%20brief%20work%20transformed%20promise%20and%20peril%20of%20ai.pdf
International
Labour Organization. (2026). New ILO–World Bank paper highlights uneven
global impact of generative AI on jobs.
https://www.ilo.org/resource/news/new-ilo%E2%80%93world-bank-paper-highlights-uneven-global-impact-generative-ai-jobs
International
Monetary Fund. (2026). Bridging skill gaps for the future: New jobs creation
in the AI age (Staff Discussion Note SDN/2026/001).
https://www.imf.org/-/media/files/publications/sdn/2026/english/sdnea2026001.pdf
International
Telecommunication Union. (2025). Measuring digital development: Facts and
figures 2025. ITU.
https://www.itu.int/itu-d/reports/statistics/facts-figures-2025/
International
Telecommunication Union. (2025). Internet use — Statistics.
https://www.itu.int/itu-d/reports/statistics/2025/10/15/ff25-internet-use/
Invest in
Estonia. (2024). Estonia ranks high in the UN E-Government Survey.
https://investinestonia.com/estonia-ranks-high-in-the-un-e-government-survey/
North, D. C.
(1990). Institutions, institutional change and economic performance.
Cambridge University Press.
OECD. (2024). How
countries are implementing the OECD principles for trustworthy AI. OECD.AI
Policy Observatory. https://oecd.ai/en/wonk/national-policies-2
OECD. (2026). OECD.AI
Policy Navigator. https://oecd.ai/en/dashboards/national
OECD &
World Bank. (2023–2025). Enterprise Surveys: Digital adoption and firm
productivity. World Bank Group.
Pandey, B.
(2026). Digital public infrastructure (UPI, Aadhaar) and entrepreneurship
growth in India: A case-based macroeconomic and structural analysis. SSRN.
https://ssrn.com/abstract=6633138
PwC. (2017). Sizing
the prize: What's the real value of AI for your business and how can you
capitalise? PwC Global Artificial Intelligence Study.
Reach Alliance.
(2026). From A to O-positive: Blood delivery via drones in Rwanda.
University of Toronto.
https://reachalliance.org/reports/ziplines-impact-on-health-outcomes-of-the-hardest-to-reach-in-rwanda/
Schultz, T. W.
(1961). Investment in human capital. American Economic Review, 51(1), 1–17.
Stanford
Institute for Human-Centered Artificial Intelligence. (2025). The 2025 AI
Index Report. Stanford University.
https://hai.stanford.edu/ai-index/2025-ai-index-report
Stanford
Institute for Human-Centered Artificial Intelligence. (2026). The 2026 AI
Index Report. Stanford University.
https://hai.stanford.edu/ai-index/2026-ai-index-report
Think Global
Health. (2026). Drones deliver humanitarian aid in Africa. Council on
Foreign Relations. https://www.thinkglobalhealth.org/article/drones-deliver-humanitarian-aid-africa
UNESCO.
(2023–2025). AI and education: Guidance for policymakers. UNESCO.
United Nations
Department of Economic and Social Affairs. (2024). UN E-Government Survey
2024. United Nations.
World Bank.
(2016). World Development Report 2016: Digital dividends. World Bank
Group.
World Bank.
(2021–2025). World Development Report series and Digital Adoption Index.
World Bank Group.
World Economic
Forum. (2025). Africa is harnessing technology to leapfrog towards growth.
https://www.weforum.org/stories/emerging-technologies/africa-leapfrog-moment-harnessing-technology-green-growth-and-regional-integration-for-global-value-chains/
World Economic
Forum. (2020). How Estonia's digital society became a lifeline during
COVID-19. https://www.weforum.org/stories/2020/07/estonia-advanced-digital-society-here-s-how-that-helped-it-during-covid-19/
World Economic
Forum. (2017). The global economy will be $16 trillion bigger by 2030 thanks
to AI. https://www.weforum.org/stories/2017/06/the-global-economy-will-be-14-bigger-in-2030-because-of-ai/
World Economic
Forum. (2024–2025). The Future of Jobs Report. World Economic Forum.
No comments:
Post a Comment