Oversight crucial as digital government blossoms

As governments digitalise, well-defined oversight of digital platforms becomes critical to ensure that they remain resilient, secure, and aligned with policy objectives.

NCS’ Government Strategic Business Group (Gov+) Chief Executive, Sam Liew, says this oversight must extend beyond traditional IT boundaries.

“There is a need for a holistic approach that encompasses compliance, digital resilience, proper security measures, and systematic access controls,” he shares.

Speaking to GovInsider, Liew, who is also the Deputy CEO of NCS Group, a digital and technology services company and part of Singtel, adds that a reliable and consistent digital platform supports not only technical performance, but can shore up public confidence in digital government initiatives.

He notes that audit and compliance are two important aspects of governance oversight that help build robust digital systems.

“Modern approaches to audit within digital government seek to embed compliance and oversight into everyday operations, rather than treating them as occasional check-ins or afterthoughts,” says Liew.

This means that compliance cannot only be maintained during the initial implementation but also throughout ongoing operations and system upgrades.

“These continuous compliance mechanisms help digital governments avoid operational drifts and ensure that new initiatives remain within regulatory and ethical boundaries,” Liew says.

Ensuring good “citizen experience”

While audit and compliance provide oversight to ensure good governance, public sector service delivery platforms must also be designed with end users in mind, “ensuring that digital government services are intuitive and accessible to all”, Liew adds.

In this context, he shares the concept of "citizen experience”.

According to him, citizen experience is a public sector analogue of the private sector concept of customer experience.

“Citizen experience expands upon traditional user experience by considering the specific and diverse needs of the public, rather than just generic users. This shift reflects a deeper commitment to public service and responsiveness,” he notes.

Governments that are successful in digital government implementation are making significant investments and efforts to improve delivery platforms, focusing on interoperability and application programming interface (API) integration.

By continuously evolving delivery platforms, governments can support the diverse and growing expectations of their stakeholders, Liew says, adding that the foundation of resilient digital government infrastructure begins with robust network connectivity.

Network segmentation required

Liew emphasises that, in both the public and private sectors, entities are moving beyond generic, unified networks to more deliberate designs that incorporate network segmentation and separation.

Depending on the sensitivity of each segment, users implement varying levels of access control and security measures like zero trust. He notes that such a design helps ensure both performance and security.

This approach recognises the growing sophistication of cyber threats and the need to prevent unauthorised access or lateral movement within digital infrastructures, Liew notes.

While moving to a hybrid multi-cloud environment helps, he adds that a thoughtful blend of public, private, and on-premises cloud solutions allows public sector organisations to tailor their approach to the varying sensitivity and criticality of different types of services and data.

“Less sensitive data may be stored in more flexible, scalable public clouds, while sensitive or high-value information remains secure in sovereign or private cloud environments.

“This hybrid approach provides the flexibility needed for innovation and service delivery, while ensuring that data protection and sovereignty requirements are met,” he adds.

Data governance in the age of AI

Data governance has become a crucial factor in the public sector as it ramps up the use of artificial intelligence (AI), especially generative AI (GenAI) and agentic AI applications.

He notes that data governance includes robust frameworks and policies that ensure AI technologies “not only comply with regulations but also uphold principles of fairness, non-discrimination, and explainability”.

According to him, the public sector needs to develop what he calls “AI-ready data”.

Liew says this is about assembling “high-quality, well-organised, and ethically managed data” that can be securely used by advanced AI models while upholding the principles of fairness, non-discrimination and explainability.

To subscribe to the GovInsider bulletin, click here. 

Privacy, stewardship, and minimal retention of personal information are fundamental, and organisations should put in place rigorous policies ensuring that both technical capabilities and ethical standards are upheld, he adds.

Liew notes that there are several early examples of agentic AI applications in the Singapore public sector.

AI-powered assistants

He shares an example of a public sector agency that has developed an AI-powered assistant to help public officers quickly access and interpret security documents, human resource (HR) policies, and other internal resources “that would otherwise be challenging to navigate”.

Another use case involves deploying GenAI and agentic functions in government call centres, with agents able to quickly summarise and document call transcripts for integration into customer relationship management (CRM) systems, he adds.

“These examples show how agentic AI and GenAI can streamline workflows, reduce manual effort, and support more informed decision making,” he adds.

Despite these advances, Liew adds a note of caution.

“The public sector’s approach to agentic AI remains careful and methodical [and] most projects are still in a pilot or experimental phase, with government leaders acutely aware of the need for validation and robust governance before granting AI systems any significant autonomy,” he adds.

Liew notes that there is [and has to be] particular emphasis on retaining a "human-in-the-loop" for oversight, especially as agentic AI moves from supporting roles to potentially handling core government decisions.

“This cautious stance is necessary due to the critical importance of accuracy, public accountability, and the ethical implications of delegating authority to autonomous systems,” he adds.

Need to build human capital

“Technological advancements alone are insufficient without continuous investment in human capital to build, maintain, and innovate digital infrastructure, Liew notes.

Liew, a former president of the Singapore Computer Society (SCS), has worked on skills development. He notes that the Republic’s approach to training and skills upgradation has evolved from sporadic, ad hoc training to a more systematic and policy-driven framework for upskilling the entire workforce.

“This constant cycle of learning and adaptation has become especially important as digital government initiatives hinge on expertise in cloud, AI, cybersecurity, and emerging technologies,” he adds.