AI is absolutely everywhere. At the IMF and World Bank Annual Meetings in Bangkok, from October 12 to 18, the technology appeared in forecasts for global growth, conversations about jobs and warnings about financial stability. Its reach has become an economic fact. The institutions responsible for managing its consequences are still finding their footing.
The RegTech joined these discussions through the participation of its chief executive, Igor Ujhazi. For a company working at the intersection of public-sector digitalization, regulation and market entry, the significance of Bangkok lay in the relationship between technological ambition and the capacity of governments to carry it through.
Ujhazi has repeatedly raised a concern that now deserves a central place in economic policy: fascination with AI’s rapid development has outpaced the legal and administrative arrangements governing its use. Governments need credible rules for making decisions, assigning responsibility and protecting the people affected by automated systems. The cost of delay is growing.
5 Key Takeaways
- Governance lags behind AI: AI’s rapid progress has outpaced the legal and administrative rules needed to govern it. Clear frameworks for accountability, liability, and citizen protection are still missing, and the cost of delay is rising.
- AI is already reshaping growth and risk: Investment in AI infrastructure is driving uneven global growth while creating new financial stability risks. Over-optimistic profit assumptions and concentrated exposures could trigger sharp market corrections that affect tax revenues and sovereign borrowing costs.
- Stronger revenue systems are essential: Rising debt-service burdens (around 14% of revenue in lower-income countries) leave little room for public investment. Better tax design, administration, and reliable transaction records are critical to creating fiscal space.
- “AI paralysis” is a policy risk: Many governments see AI’s potential but cannot move from interest to responsible use because legal authority and safeguards remain incomplete. The result is stalled projects or unsupervised adoption that erodes trust.
- Responsible AI needs strong institutions: Clear laws, interoperable registries, auditable data, and resilient systems are the foundation. Technology only delivers public value when governments can lawfully control, audit, and correct it.
An Economy Living on Diminishing Reserves
The IMF’s October World Economic Outlook projected global growth of 3.1% in 2026 and 3.4% in 2027. Inflation was expected to reach 4.7% this year before easing to 4.1% next year. These figures describe an economy that continues to expand under considerable pressure from energy disruption, trade uncertainty and expensive borrowing.
The war in the Middle East has raised energy costs and strained supply routes. Meanwhile, investment in AI infrastructure has supported demand for semiconductors, data centres and power generation. The resulting growth is uneven. Economies supplying the technology have benefited disproportionately, while many energy importers are confronting another squeeze on household incomes and public budgets.
Bangkok’s practical lesson was that resilience has limits. Emergency buffers are being drawn down, bond markets are increasingly sensitive to government deficits, and public investment must compete with debt servicing. Ministries of finance face difficult choices over which programmes to fund and which reforms can produce durable economic returns.
AI is Everywhere: Investment Is Redrawing the Map of Growth
AI is ever present in the investment cycle as well. Capital is flowing into computing infrastructure and electricity supply well before the full productivity gains can be measured. For countries embedded in technology supply chains, that spending is already supporting output, exports and employment. Elsewhere, the opportunity depends on affordable power, connectivity, skills and access to reliable systems.
The IMF’s Global Financial Stability Report warned that optimistic assumptions about AI profits could produce sharp market corrections if returns disappoint. Concentrated exposures, borrowing and links between technology firms and financial markets make that prospect relevant well beyond Silicon Valley. Cyberattacks present a further systemic risk as finance becomes more connected.
Governments have a stake in the outcome even when they invest little directly in AI. Market adjustments affect tax revenues, pension assets and the cost of sovereign borrowing. Changes in employment reshape the tax base. AI is Everywhere in economic policy, from changes in labour demand to the exposure of sovereign finances. Governments need to assess these effects as adoption spreads.
Fiscal Credibility Begins with the Ability to Collect
Debt was an unavoidable part of the meetings. In lower-income countries, average debt-service payments have risen to about 14% of government revenue, according to figures discussed in Bangkok. Every additional payment narrows the funds available for schools, health systems, infrastructure and adaptation to future shocks.
For The RegTech, these pressures strengthen the case for domestic revenue mobilization. The IMF’s October Fiscal Monitor, titled Taxing Better to Boost Growth, emphazised sounder VAT and corporate tax design, stronger administration and clearer legal frameworks. Better collection and simpler compliance can create room for investment without depending entirely on higher statutory tax rates.
That distinction matters in practice. Tax authorities need to know which economic activities are taking place, which transactions support declared liabilities and where compliance risks are concentrated. Transaction-level evidence, reliable taxpayer identification and connected administrative systems can improve oversight. Their value comes from the integrity of the records and the laws that make those records usable.
Development Finance Has an Institutional Test
World Bank President Ajay Banga used the annual plenary to argue for bringing public finance, private investment, philanthropy and technology together around national priorities. He also set out an employment challenge: roughly 1.2 billion young people are expected to enter labour markets over the next 10 to 15 years, against an estimated 400 million new jobs.
Those figures give urgency to the search for productive investment. Yet capital needs working institutions. Investors assess legal certainty, the quality of public administration, payment systems, regulatory predictability and the capacity of governments to honour commitments. Weakness in any of these areas raises costs, especially in economies with limited fiscal buffers.
Accordingly, The RegTech’s experience in public-sector technology reinforces a straightforward proposition: economic modernization is sustained by institutions capable of operating and enforcing it. Digital identity, dependable registries, transparent procurement and trusted transaction records can make reform easier to finance and easier for citizens and businesses to use.
AI is Everywhere: Paralysis is Becoming a Public-Policy Risk
For many administrations, the next step remains uncertain. They recognise the potential for faster services, stronger fraud detection and better resource allocation. They also face unanswered questions about lawful data use, discrimination, liability, procurement and the appeal of automated decisions.
This is what we describe as AI paralysis: the inability to move responsibly from interest to implementation because institutional authority and safeguards are incomplete. It can appear as postponed projects, uncoordinated pilot schemes or the quiet adoption of tools without adequate supervision. Each outcome carries costs for public confidence and administrative effectiveness.
The pace of innovation makes the problem unusually demanding. Models evolve faster than ordinary public procurement cycles, while data may cross borders and decisions may depend on systems that governments do not control. Legal uncertainty leaves civil servants unsure about what they may approve, and citizens unsure about where accountability rests.
At an IMF discussion in Bangkok, managing director Kristalina Georgieva compared the speed of AI development to a racing car that needs protective barriers. The image captured the issue confronting policymakers. Practical safeguards must become part of deployment decisions from the outset, with obligations that remain enforceable as technology changes.
Turning Governance Principles into Usable Law
There has been international progress. The UN’s Global Digital Compact and its scientific and policy mechanisms have expanded cooperation on AI governance. Governments still need to translate broad principles into operational rules suited to their own institutions, laws and administrative capacities.
In public administration, those rules must establish who may deploy an AI system, which decisions require human approval, how performance is tested and how an affected person can challenge an error. They must also address privacy, cybersecurity, data retention, vendor access and the documentation needed for independent audit. A procurement contract cannot substitute for statutory authority.
Igor Ujhazi’s recurring argument is that legal certainty can enable responsible innovation. Clear mandates reduce the hesitation of officials, establish duties for suppliers and give the public a basis for trust. Governments that build this foundation can pursue valuable applications with greater confidence and correct failures without leaving accountability unresolved.
The Infrastructure Beneath Responsible AI
Rules require dependable information systems. A public agency relying on inconsistent registries, incomplete transaction records or weak identity verification will struggle to establish whether an automated recommendation is accurate. Moreover, the same weaknesses undermine conventional administration, making digital reform a broader institutional priority.
The RegTech sees particular value in interoperable registries, verified credentials, auditable data exchange and operational resilience. In fiscal administration, an invoice linked to an identifiable taxpayer and preserved as reliable evidence offers a stronger basis for analysis than an isolated spreadsheet or unverified submission. AI can then assist officials working within a traceable decision process.
Resilience also deserves attention. Network failures, cyber incidents and service interruptions are ordinary planning concerns for national systems. Human review, secure records, continuity procedures and tested recovery arrangements belong in the architecture before automated processes are entrusted with consequential public functions.
Final Verdict: Make Technology Governable!
The meetings in Bangkok brought fiscal pressure, technological investment and development finance into the same conversation. Their combined implication is clear for governments trying to modernize with limited resources. Public institutions will need to make better decisions using stronger evidence, while preserving lawful authority over the tools they adopt.
AI is everywhere, and in the words of Kristalina Georgivea, “love it or hate it,,, it’s here to stay”. While its development is genuinely fascinating, the economic opportunity around it deserves serious attention. The RegTech believes the next phase depends on treating legislation, accountability, institutional skills and reliable digital infrastructure as essential parts of that opportunity. Governance determines how far innovation can safely travel through the public sector.
For us, participation in the 2026 Annual Meetings reinforced a conviction that extends across fiscal reform and wider digital government: technological capability acquires public value through institutions people can trust. Countries that clarify the rules, strengthen their evidence and invest in administrative competence will be better prepared to use AI where it serves a demonstrable purpose. Bangkok offered an assessment of the pressures ahead. The work now belongs to governments, regulators and us, technology partners. Responsible adoption will be measured in lawful decisions, credible records, effective services and the confidence of the people who depend on them.
We are here to help governments, financial institutions, and businesses to effectively comply with growing regulatory requirements through technology.
