…the lessons of 2025 indicate that AI’s impact in 2026 will depend not on deployment speed, but on governance readiness — including data quality, interoperability, accountability, and safeguards for trust. These emerging trends, and the choices governments face as AI adoption deepens, will be explored in greater detail in the coming weeks.
In 2025, technology did more than digitise public services — it revealed how governance actually functions. Across countries like the United States, Nigeria, Estonia, India, Brazil and Singapore, digital systems became a mirror through which institutional coherence, accountability, and capacity were made visible. Where governments had clear rules, interoperable systems, and internal delivery capability, technology amplified effectiveness. Where authority was fragmented and execution weak, digital tools exposed — and often intensified — existing governance failures.
This was particularly evident as artificial intelligence (AI) moved from experimentation to early adoption across public sectors. In 2025, governments were forced to confront questions of data quality, institutional readiness, accountability, and trust. Those with strong digital foundations used AI to improve productivity, planning, and service responsiveness. Those without them discovered that AI magnifies structural weaknesses — poor data, unclear ownership, and limited capacity — rather than resolving them. In this sense, digital systems and AI together showed that technology does not fix governance; it reveals it.
In Nigeria, 2025 renewed debate around local government autonomy alongside growing interest in digitisation and AI-enabled public services, but it also exposed a persistent gap between ambition and execution. Many core government functions — civil registration, land administration, budgeting, and procurement — remain manual and siloed, particularly at the local level, while discussions about AI adoption have largely stayed within policy documents rather than practice, due to limited interoperability, inconsistent data standards, and uneven digital capacity. Digital transformation in Nigeria cannot be reduced to websites or portals; it requires a deliberate shift toward modular, interoperable systems that enable coordination across tiers of government and make service delivery measurable and accountable. The central governance lesson from 2025 is therefore clear: without shared systems, reliable data, and delivery capability at the subnational level, neither digitisation nor AI can scale meaningfully, and autonomy risks decentralising inefficiency, rather than improving service delivery.
In the United States, 2025 saw the increased deployment of AI tools across federal, state, and local governments — particularly for administrative processing, fraud detection, and service triage. However, these efforts exposed long-standing governance challenges. AI systems often had to operate on top of fragmented legacy infrastructure, uneven data quality, and siloed authority across agencies. Where agencies had modular systems, shared platforms, and in‑house digital teams — such as those supported by the US Digital Service (USDS) — AI tools could be integrated more quickly into existing workflows, thereby improving turnaround times and decision support. Where they did not, AI adoption was slower, more contested, or constrained by risk concerns. For example, the US Government Accountability Office’s reviews of the Department of Defense’ IT and business‑systems modernisation point to fragmented legacy environments, inconsistent data practices, and significant cybersecurity requirements that complicate efforts to move AI capabilities beyond pilots and prototypes into day‑to‑day operations. The lesson was clear: AI does not bypass governance complexity; it intensifies the need for coherent architecture, clear accountability, and institutional coordination.
Globally, 2025 highlighted that countries making progress with AI and advanced digital government had made deliberate governance choices early. Estonia used its interoperable data exchange platform, X-Road, to support secure automation and data sharing across public services without duplicating records. India built on its digital public infrastructure — anchored by Aadhaar and UPI — to apply AI to service delivery and administration at population scale. Brazil leveraged its publicly governed instant payment system, Pix, to embed automation into financial services while expanding inclusion and reducing reliance on cash. Singapore, through LifeSG, integrated AI and digital tools into life-event-based service journeys, allowing citizens to access multiple services through a single, coordinated workflow. Across these contexts, the pattern was consistent: AI delivered value where governance frameworks, shared data standards, and institutional capacity were already in place.
|
|
What Technology Taught Us About Governance in 2025
Across contexts, several governance lessons emerged.
- Interoperability Is a Governance Decision: Interoperability is often framed as a technical challenge. In reality, it is a governance choice. Governments that failed to mandate shared standards created digital silos that undermined planning, targeting, and oversight. Those that enforced interoperability enabled coordination across institutions and tiers of government.
- Inclusion Reflects Institutional Priorities: Systems designed only for high bandwidth and high digital literacy excluded large segments of the population. In contrast, governments that prioritised multi-channel access — USSD, messaging platforms, and assisted service points — were better able to reach rural and low-income populations. Inclusion is not a feature; it is a policy decision.
- Data Protection Is Now Operational Governance: In 2025, data protection frameworks moved decisively from policy to practice. Enforcement actions, breach disclosures, and public scrutiny made clear that digital delivery without credible data governance erodes legitimacy. Trust became a prerequisite for adoption, not a communications strategy.
- Capacity Inside Government Determines Sustainability: Technology performed best where governments invested in internal delivery capability — product managers, service designers, engineers, data analysts, and change managers embedded within institutions. Where transformation was entirely outsourced, systems degraded once contracts ended or leadership changed.
- Citizen Engagement Should be Continuous, Not Episodic: In 2025, technology enabled continuous feedback loops — using data on service usage, grievances, and public sentiment — allowing governments to shift from static policy cycles to more iterative and adaptive decision-making. Governance systems that used technology to learn and adjust, rather than simply automate, proved more responsive and resilient.
- Technology Is a Test of Institutional Readiness: What 2025 ultimately revealed is that technology is not the binding constraint. The constraint is institutional readiness: the ability to coordinate across agencies, fund delivery consistently, protect citizens’ data, and measure outcomes. Governments that struggled were not those without platforms, but those with too many disconnected ones.
Technology will not fix governance, but it will continue to reveal it. In 2025, digital systems made visible what had long existed beneath the surface: How governments coordinate authority, prioritise resources, and deliver public value under constraint. Digital governance will not be built in a day. But the work can — and must — begin deliberately. The governments that act on the lessons of 2025 will not be those with the most ambitious rhetoric, but those willing to invest in systems, capacity, and accountability. Technology has already taught us what matters. The remaining question is whether governments are prepared to respond.
Looking ahead, 2026 is likely to see the accelerated adoption of AI across governments, driven less by novelty and more by rising citizen expectations and operational pressure. Global evidence already points in this direction: an OECD survey finds that a growing share of citizens expect faster, more personalised, and more responsive public services, comparable to private-sector experiences. At the same time, the World Bank notes that client governments are increasingly seeking support on AI as part of broader digital transformation programmes aimed at increasing efficiency and quality of service delivery, improving citizen engagement, and modernising core operations.
Importantly, this shift is being reinforced by more targeted public and philanthropic funding for GovTech and AI solutions. In the United States, for example, states such as New Jersey are channeling new resources into statewide IT modernisation, AI innovation challenges, and cross‑government AI tools, signalling a gradual move from small pilots toward institutionalised digital and AI‑enabled operations. Industry research further suggests momentum is building as McKinsey estimates that generative AI could unlock significant productivity gains in the public sector, especially in administrative and citizen-facing functions. However, the lessons of 2025 indicate that AI’s impact in 2026 will depend not on deployment speed, but on governance readiness — including data quality, interoperability, accountability, and safeguards for trust. These emerging trends, and the choices governments face as AI adoption deepens, will be explored in greater detail in the coming weeks.
Shona 0. Oluwatola is a public policy and AI strategist with expertise in technology, public health, and governance. She is the Founder of AccessGov, a GovTech company driving digital transformation in government, and co-founder of CDIAL AI, which builds inclusive speech and data technologies for African languages. Through her work with Helium Health, she advances digital innovation in healthcare across Africa. A Doctoral Fellow at the RAND Corporation, her research focuses on digital governance and policy innovation in low- and middle-income countries.
