Why incomplete data shouldn’t hold back AI innovation in the public sector

Applications of AI


Southeast Asia dominates the adoption of artificial intelligence (AI) globally, piloting and expanding more AI projects than the United States or the Asia-Pacific region (excluding China and India), according to a recent McKinsey study.

The study highlights that reliable data flows are a key driver in unlocking the region’s AI potential.

For the region’s public sector, the risks of getting AI right are even higher, as governments have a responsibility to responsibly provide essential services while protecting citizens’ data.

Across ASEAN, the adoption of AI in the public sector is already taking shape through real-world use cases rather than abstract experiments.

Gartner, the world’s leading research and advisory firm, points to emerging applications such as AI-assisted public response services, incident handling, and maritime and transportation operations.

This is paralleled by the use of digital innovation labs and controlled “data sandbox” environments that allow agencies to test AI (including synthetic data) while protecting sensitive information.

These examples highlight that ASEAN’s public sector digital systems are already operational at national scale, supporting identity, mobility, security, and regulatory functions for millions of citizens.

Rather than waiting for perfect data before implementing AI, NEC recommends that agencies start with clearly defined business goals and use cases.

Organizations looking to start implementing AI, whether private companies or government agencies, will remain in “pilot mode” if they want perfect conditions before deploying AI, NEC said.

The organization sees the risk of waiting for perfect data as slowing value creation and learning for these institutions.

In this context, technology partners working with governments need a deep understanding of public sector operations, including regulatory, safety, and service delivery realities.

NEC has leveraged its long-standing collaboration with ASEAN governments to support public sector initiatives.

These efforts provide experience in operating trusted digital systems at scale, ranging from disaster prevention systems to resident services, including securing more than 100 million biometric IDs across the region.

“A key consideration is whether the available data is sufficient to make informed decisions, not just to achieve statistical perfection,” an NEC spokesperson said.

NEC offers one or both of two types, black box AI and white box AI, depending on the use case. Image: NEC’s AI guidebook

NEC’s AI Guidebook emphasizes the value of human judgment and governance in AI system development, ensuring that these systems can operate with incomplete and evolving data.

Where is AI in the public sector heading?

Industry forecasts indicate that public sector AI investment will continue to grow, reflecting a broader shift to the use of digital technologies and underlying infrastructure to support productivity, service continuity, and economic resilience.

Rather than deploying AI as a standalone tool, the new direction is to embed it into trusted and secure platforms that unify data, operations, and governance.

This evolution is increasingly seen as critical to achieving long-term economic and social outcomes, particularly as governments balance innovation with accountability and public trust.

How can agencies move forward?

Start with the problem, not the data.

Rather than attempting broad system change, public organizations can start by identifying high-impact, operationally meaningful decisions.

“In practice, this means agreeing in advance what decisions the AI ​​will support, how humans will be held accountable, and what evidence, particularly data and reasoning, must be available to justify outcomes,” says an NEC spokesperson.

Following initial pilots and moving into production deployments, the next step is for these leaders to consider incorporating AI into their workflows.

“As AI adoption expands, organizations often struggle to explain why AI produced results, especially when decisions impact people,” the spokesperson says.

For this reason, NEC is promoting an explainable AI approach, also known as white-box AI, that can tell leaders the reason behind an output.

This ensures accountability even when data is fragmented.

NEC is tackling the LLM hallucination problem by providing three functions: Quality Checker, LLM Explainer, and Fact Checker to verify LLM output and improve safety. Image: NEC

A practical approach to explainable AI

NEC has developed three functions: “Quality Checker,” “LLM Explainer,” and “Fact Checker” to support the explainability, traceability, and transparency that are increasingly required in the public sector.

These capabilities were recognized at the Newsweek AI Impact APAC & EMA Awards, which spotlight organizations that are innovatively using AI to solve everyday challenges.

By deploying emerging technologies such as agent AI, NEC will take a zero-client approach and serve as the first client to deliver informative real-world experiences to public sector customers.

To further increase confidence in the output and enable proactive action and quick decision-making, NEC uses management dashboards to visualize performance data and incorporate predictions based on the current performance of the AI ​​system.

NEC highlights key lessons for organizations, including embedding AI into daily workflows, maintaining human accountability, and ensuring transparency throughout the AI ​​lifecycle. These are also relevant to public sector leaders.

NEC is creating the future of the public sector

Going forward, the imperative for public sector organizations is to move beyond isolated pilots to scalable, reliable AI infrastructure that can support nation-wide services and long-term societal outcomes.

This change is already underway across the region. A joint study by the Economic Development Board of Singapore (EDB) and McKinsey finds that organizations in Southeast Asia are moving from experimentation to full AI implementation, reflecting growing confidence in AI’s role as part of core digital infrastructure rather than standalone trials.

When applied intelligently and thoughtfully, AI can help enhance decision-making across areas such as identity, mobility, safety, and citizen services while maintaining accountability and public trust.

For NEC, this initiative is about aligning innovation with governance, transparency and real operational needs, and it’s an exciting time to work with the public sector to create the future.



Source link