Resisting the Allure of "Artificial": A Strategic Framework for Sovereign Technology in Emerging Economies
Executive Summary
Emerging and frontier markets are at a critical juncture, facing immense pressure to adopt and develop sovereign technologies, particularly in the realms of central bank digital currencies (CBDCs) and national artificial intelligence (AI) capabilities. Driven by the promise of technological leapfrogging and geopolitical positioning, many governments are launching ambitious, high-profile projects. However, a closer examination reveals a perilous trend: a rush to innovate without a clear strategy, resulting in costly failures that drain scarce public resources and erode public trust. This brief argues that the pursuit of technological sovereignty, while a valid long-term goal, is being dangerously conflated with the premature launch of vanity projects.
Drawing on data from large-scale IT project failures and specific case studies of struggling CBDC and AI initiatives, this analysis presents a more rigorous, data-driven framework for policymakers. It moves beyond generic advice to offer concrete, risk-weighted policy recommendations designed to help emerging economies distinguish between strategically sound investments and the siren call of technological prestige. The central message is not to abandon ambition, but to ground it in a disciplined, evidence-based approach that prioritizes foundational capacity and genuine user needs over the allure of the artificial.
The Sobering Reality of Public Sector Technology Projects
The ambition to launch transformative national technology projects often collides with a harsh reality: public sector IT initiatives have a historically high rate of failure. Before an emerging economy even considers a project as complex as a CBDC or a sovereign AI model, it must internalize the sobering statistics that govern such endeavors. Research from McKinsey provides a stark quantitative baseline, revealing that public-sector IT projects are systematically more prone to failure than their private-sector counterparts. A staggering 80% of government IT projects overrun their schedules, and nearly half exceed their budgets. The average cost overrun for public projects is nearly three times higher than in the private sector, with 14% of government projects doubling their initial budget.
These are not minor deviations; they represent a systemic misallocation of capital and human resources. The data shows an exponential increase in cost overruns correlated with project duration, a particularly damning fact given the long-term nature of building and maintaining sovereign tech stacks. The reasons for this are manifold and deeply embedded in the nature of public administration: complex stakeholder landscapes, ineffective risk management, chronic difficulties in attracting and retaining top technical talent, and slow governance processes. This challenging environment forms the backdrop against which the high-stakes bets on CBDCs and sovereign AI are being placed. The question for policymakers is not simply whether a technology is promising, but whether their government is foundationally equipped to defy these overwhelming odds.
Case Studies in Failure: The Global CBDC Experiment
The global experiment with central bank digital currencies provides a powerful, contemporary set of case studies in how even well-intentioned technological projects can fail to gain traction. Despite 137 countries exploring a CBDC as of mid-2025, the few that have fully launched have seen adoption rates that can only be described as catastrophic failures . These are not early-stage hiccups; they are market rejections.
Nigeria’s eNaira, launched in 2021, is perhaps the most salient example. In a nation with a vibrant, youthful population that has enthusiastically adopted private cryptocurrencies (over 50% of the population), the official eNaira has been met with near-total indifference. By 2022, the adoption rate was a mere 0.5% . The government’s attempt to force adoption by imposing severe cash withdrawal limits backfired spectacularly, leading to public protests and a thriving black market for physical currency. The project failed because it did not solve a problem that a plethora of existing mobile money solutions had not already addressed more effectively and with greater user trust.
This is not an isolated incident. The Bahamas’ Sand Dollar, one of the world’s first live CBDCs, has an adoption rate of just 0.08%, with only about 300,000 of the 386 million Bahamian dollars in circulation being digital . Even in China, where the state has immense power to drive adoption, the e-CNY pilot, while processing large transaction volumes, has struggled to achieve meaningful, active use among its 261 million registered users, with most remaining inactive . The lesson is clear: if a new technology does not offer a tangible, significant improvement in convenience, cost, or accessibility over existing, trusted alternatives, the public will not use it, regardless of government pressure.
Sovereign AI: A Bridge Too Far?
The challenges facing sovereign AI initiatives are even more profound. While the allure of national AI models is strong (competitiveness, sovereignty, and retained influence) the foundational requirements are immense. The United Nations Development Programme (UNDP) warns of a “Next Great Divergence,” where the gap between AI-ready and unready nations widens dramatically . The central fault line, according to the UNDP, is capability. This encompasses three critical areas where most emerging markets are critically deficient:
- Computing Power: The development of large-scale AI models requires massive, centralized computing resources that are currently concentrated in the hands of a few global hyperscalers. Emerging economies face a significant “compute gap,” lacking the data centers and specialized hardware to train and operate these models at a competitive scale.
- Human Capital: There is a global shortage of top-tier AI talent. Emerging markets struggle to compete with the salaries and research opportunities offered in developed AI hubs, leading to a persistent brain drain.
- Data Infrastructure: High-quality, well-structured, and comprehensive datasets are the lifeblood of AI. Many emerging economies lack the robust data infrastructure and governance frameworks necessary to create and leverage these assets effectively.
Without these foundational pillars, a sovereign AI project risks becoming a “Potemkin AI”—a facade of technological prowess with no real substance behind it. It becomes a sinkhole for public funds, diverting investment from more achievable and impactful goals, such as investing in digital literacy, strengthening data governance, and building targeted AI applications for specific public services, as seen in successful, smaller-scale projects in Bangkok and Singapore .
A Risk-Weighted Framework for Sovereign Technology
To avoid these pitfalls, policymakers must move beyond simple cost-benefit analysis and adopt a risk-weighted evaluation framework. This involves not only assessing the potential rewards but also honestly appraising the probability of failure and the magnitude of the potential losses. The following four-gate framework is proposed as a disciplined pathway for decision-making.
| **Stage** | **Key Question** | **Subsequent Action** |
|---|---|---|
| #### Problem-Solution Fit Assessment | Does this project solve a high-priority, validated problem for a large segment of the population that is not already being adequately addressed by the private sector? | Commission independent, user-centric research to validate the problem. Conduct a thorough market analysis of existing solutions. If a clear, unaddressed need is not identified, the project should not proceed. |
| #### Foundational Capability Audit | Does the state possess the requisite in-house technical talent, data infrastructure, and regulatory agility to execute this project effectively? | Conduct a comprehensive audit of internal capabilities. Identify gaps in talent, infrastructure, and regulatory frameworks. If critical deficiencies cannot be realistically addressed, the project should not proceed. |
| #### Pilot and Iteration | Can the project be launched as a small-scale, measurable pilot with clear success metrics and a mechanism for rapid iteration and failure? | Design a pilot program with a limited scope and budget. Establish clear KPIs for success and failure. Implement a feedback loop for continuous improvement. Only scale if the pilot demonstrates clear, measurable success. |
| #### Scalability and Sustainability | Is there a clear, viable path to scale the project nationally and ensure its long-term sustainability without continuous, disproportionate public subsidy? | Develop a detailed scaling strategy, including funding models, operational plans, and governance structures. Assess the long-term economic and social impact. If the project cannot demonstrate a path to sustainable scale, it should be re-evaluated or terminated. |
Conclusion
The pursuit of technological sovereignty in emerging economies is a complex and often perilous endeavor. The allure of cutting-edge technologies like CBDCs and AI can be powerful, but the sobering reality of public sector IT project failures demands a more disciplined, risk-weighted approach. By adopting a rigorous framework that prioritizes problem-solution fit, foundational capability, iterative piloting, and sustainable scalability, policymakers can navigate the treacherous waters of technological innovation. The goal is not to resist technology itself, but to resist the allure of the artificial—the superficial prestige of vanity projects—in favor of genuine, impactful, and sustainable development. Only then can emerging economies truly harness the transformative power of technology to build a more prosperous and equitable future.