Douglas Levermore | The Next Frontier of Smarter Government

Every budget season forces governments to answer the same difficult question: which investments deserve scarce public dollars? 

This year, however, that question has acquired an entirely new dimension. Artificial intelligence (AI) is no longer simply another area of technology policy. It is becoming both a major public investment in its own right and one of the most powerful tools governments have ever possessed to determine what they should invest in next.

As governments around the world move through their annual budget cycles, finance ministries and planning commissions find themselves confronting one of the most consequential exercises in public administration. Capital budgets are finalised, project pipelines are scrutinised, and difficult decisions are made about which roads, hospitals, schools, water systems, and other public investments will receive scarce resources. It is during these periods that public investment management (PIM) becomes most visible, providing the discipline to screen projects, conduct rigorous appraisals, prioritise competing proposals, and align investment decisions with national development goals and fiscal realities. Anyone who has worked inside a finance ministry understands that every project approved for funding represents dozens that never make it onto the capital programme. Public investment has always been an exercise in disciplined choice rather than unlimited ambition.

This year’s budget discussions, however, feel markedly different. AI has moved from being an emerging technology to becoming a strategic consideration in national planning. Governments are no longer debating whether AI deserves public investment. They are asking a far more difficult question: how should AI-related investments be evaluated, prioritised, and governed alongside more traditional infrastructure projects? National cloud platforms, hyperscale data centres, digital-twin technologies, and predictive analytics systems are now competing with highways, hospitals, ports, and power plants for finite public resources.

The conversation has therefore shifted beyond simply adopting AI. It is rapidly becoming both a capital investment in its own right and a powerful decision-support capability capable of transforming how governments identify, appraise, implement, and monitor every other public investment.

Around the world, this shift is already taking shape. Australia’s National AI Plan envisions substantial investment in data centres, digital capabilities, and AI applications across sectors ranging from healthcare to transportation. The European Union’s InvestAI initiative aims to mobilise approximately €200 billion in AI-related investments, including advanced computing infrastructure. In the United States, federal policy has increasingly treated AI data centres as strategic national infrastructure worthy of expedited permitting and financing. While these initiatives are often presented as technology policy, they are equally exercises in public investment management.

 Governments must still answer the timeless questions of which projects deserve funding, how investments should be sequenced, and whether they will ultimately deliver value for citizens.

Perhaps the most significant transformation is occurring at the front end of the investment cycle. Traditionally, identifying investment needs required months of data collection, feasibility studies, and expert consultations. AI now allows governments to combine satellite imagery, demographic trends, mobility patterns, climate models, and economic forecasts to identify where investments are likely to generate the greatest social and economic returns. Research from the OECD suggests that AI can accelerate nearly every stage of policy development—from defining problems to evaluating alternative solutions—provided governments maintain strong standards for data quality and governance. Tasks that once demanded weeks of manual analysis can increasingly be completed in hours, allowing policymakers to spend less time assembling information and more time exercising judgment.

Project appraisal, long regarded as the cornerstone of sound public investment management, is also entering a new era. Machine-learning models are increasingly capable of supporting traditional cost-benefit analysis by forecasting demand, testing assumptions, and identifying patterns of optimism bias that have plagued infrastructure projects for decades. Traffic forecasts, revenue estimates, operating costs, and maintenance requirements can now be assessed using datasets far beyond the capabilities of conventional spreadsheet models, strengthening rather than replacing professional analysis.

Equally important, AI offers governments an opportunity to reduce one of the most persistent weaknesses in public investment systems: human bias. Too often, projects receive priority because they are politically attractive, geographically convenient, or championed by influential stakeholders rather than because they generate the greatest public value. Properly designed AI-assisted appraisal systems can standardise evaluation criteria, compare projects against historical performance, and identify proposals with the strongest long-term economic and social returns. While no algorithm should replace professional judgement or democratic decision-making, AI can provide a more objective starting point for conversations that have historically been vulnerable to political influence.

AI does not stop once projects receive approval. Throughout implementation, AI-enabled project management systems can integrate satellite imagery, sensor data, contractor reports, procurement information, and financial management systems to provide governments with near real-time visibility into project performance. Cost overruns, schedule delays, procurement anomalies, and quality concerns can be detected far earlier than under traditional reporting systems, allowing ministries to intervene before relatively small problems evolve into expensive failures. Portfolio managers, in turn, gain a clearer understanding of how individual projects are performing and where scarce resources may need to be redirected.

Ironically, AI is not only changing how governments manage investments; it is also reshaping what governments must invest in. The rapid expansion of AI is creating unprecedented demand for data centres, electricity generation, transmission networks, fibre-optic infrastructure, cybersecurity systems, and digital skills development. Much of this infrastructure will require either direct public investment or significant public support through regulation, financing, or risk-sharing arrangements. Public investment managers therefore face a dual responsibility: overseeing investments in AI infrastructure while simultaneously using AI to improve the quality of investment decisions across the broader public sector.

There is, however, a growing risk of what might be called “AI theatre”—investing in AI because it is fashionable rather than because it addresses a clearly defined public problem. Governments have witnessed technological enthusiasm before, and history shows that expensive digital initiatives can fail just as spectacularly as poorly conceived roads, bridges, or public buildings. Like every other capital investment, AI projects should be justified by demonstrable public value rather than technological excitement. Novelty has never been an adequate substitute for rigorous appraisal.

Yet technological capability alone will not guarantee better outcomes. AI is only as good as the data, governance, and institutions that support it. Poor-quality data will produce poor-quality recommendations. Weak governance can allow algorithms to reinforce existing inequalities rather than correct them. Historical investment data reflecting decades of political favouritism or regional imbalance may simply teach AI systems to replicate those same patterns unless deliberate safeguards are introduced.

This concern is particularly relevant for developing countries. The World Bank has consistently observed that weak public investment management systems erode a substantial share of the value generated by public investment through poor project selection, implementation delays, and cost overruns. In countries where appraisal standards remain inconsistent, procurement systems are still maturing, or project monitoring is fragmented, there is a temptation to view AI as a shortcut to better investment decisions. It is not. AI is not a substitute for institutional maturity. Countries cannot algorithm their way out of weak governance. AI magnifies whatever system it enters. If that system is disciplined, transparent, and evidence-based, AI becomes a force multiplier. If it is fragmented, politicised, or poorly governed, AI simply enables bad decisions to be made faster.

That is why the future lies not in choosing between AI and public investment management but in integrating the two. Public investment management provides the governance architecture that determines how projects are selected, how risks are managed, and how public resources are safeguarded. AI provides analytical capabilities that allow governments to process more information, identify emerging risks earlier, and evaluate competing investment options with greater speed and precision. Together, they have the potential to create investment systems that are not only more efficient but also more transparent, resilient, and accountable.

The history of public investment has always been a history of difficult choices. Governments decide which bridges to build, which hospitals to modernise, which schools to expand, and increasingly, which digital infrastructure will shape their economies for generations to come. AI will not eliminate those choices. It will simply make them faster and, potentially, far more consequential.

That is why the debate has never really been about technology. It has always been about governance. The countries that gain the greatest advantage from AI will not necessarily be those with the largest technology budgets. They will be those with the strongest institutions, the clearest investment disciplines, and the courage to allow evidence—not political expediency—to guide public investment decisions.

The next frontier of smarter government is not AI itself. It is an intelligent government.

The countries that lead in the decades ahead will not simply be those that purchase the fastest processors, build the largest data centres, or deploy the most sophisticated algorithms. They will be the ones that combine technological innovation with disciplined public investment management, sound institutions, rigorous analysis, and accountable leadership. AI can process vast amounts of information in seconds, but it cannot determine a nation’s values, weigh political trade-offs, or define the public interest. Those responsibilities will always belong to people.

Smarter government has never been about replacing human judgment. It has always been about improving it. If AI helps governments ask better questions, make better investment decisions, deliver projects more efficiently, and improve the lives of their citizens, then it will have fulfilled its greatest promise.

The next frontier of smarter government is not machines that think like people. It is governments that make wiser decisions because they have learned how to use machines wisely.

---

Douglas Levermore, MBA, JP, is an independent management consultant and the founding Executive Director of Jamaica’s Public Investment Management Secretariat (PIMSEC)—the government unit established to strengthen project appraisal, fiscal discipline, and oversight of public investment, now known as the Public Investment Appraisal Branch (PIAB) within the Ministry of Finance and the Public Service. He also serves as a FINRA arbitrator and a commissioned Notary Public in the Commonwealth of Virginia. With experience advising governments, international development partners, public bodies, and private-sector organisations on governance, public investment management, organisational performance, and strategic reform, Douglas brings a practical, results-oriented perspective to his writing on social issues, leadership, management lessons, and organisational strategy. He is available for select international consulting, advisory, keynote speaking, and project-based engagements and may be contacted at [email protected].