In Singapore’s trust economy, private enterprises must rethink digital trust and safety

Trust is not a feature, it is a foundation that must be designed from the ground up.

Singapore has steadily built a reputation for progressiveness, public efficiency, and security. 

One area where this is particularly evident is in the trust citizens place in its digital services, built from a combination of  thoughtful strategy, robust regulations, and secure, user-centred systems. This long-term and sustainable approach has earned Singapore a top-three spot in the 2024 UN e-Government Index.

Citizens prefer to interact with government platforms that are not just inclusive, but intuitively designed to anticipate their needs, from digital identity and licensing, to public health and housing services. As citizens become increasingly accustomed to this high level of assurance, responsiveness, and transparency, businesses in the country are also expected to uphold the same level of digital trust for their customers. 

Fostering an environment to build digital trust
Singapore’s approach to digital governance offers a useful reference point for private businesses to raise the bar on digital trust. Whilst the public sector has taken a coordinated, multi-agency approach, private organisations are still finding their footing and often balancing the need for compliance with evolving customer expectations.

Singaporean organisations are aware of the risks. According to a report on trust and safety trends for 2025, 43% of local businesses plan to significantly increase investment in fraud detection in the next year. 

However, the momentum is not uniform across all facets. Just over half say they will only somewhat increase ID verification (54%), and 38% say the same for content moderation. A third (38%) are keeping Know Your Customer (KYC) spending flat.

The reasons for the investments or lack thereof are familiar: cost, technical complexity, and a lack of internal expertise. However, in today’s competitive landscape, not doing so can quickly become a liability that irreparably widens the trust gap between customers and businesses. 

Private businesses often rely on automated or outdated systems and processes that do not adequately address or resolve their issues, and can feel impersonal and opaque. In the absence of clarity or proper recourse, trust can quickly erode.

Why human-in-the-loop matters when combatting bias
Much of this erosion can be traced back to over-automation. Artificial intelligence (AI) is foundational in fraud prevention and verification. However, AI alone cannot address every challenge. Whilst it enables scale, speed, and 24/7 operations, without human oversight, it can also introduce new risks to the business, particularly around context and bias.

Take Singapore’s fraud landscape: scams here are increasingly tailored to local behaviours and cultural nuances. This means that AI algorithms that have been trained on non-geo-specific datasets can actually miss out on critical details necessary for an automated system to detect wrongdoing. 

In such cases, human-in-the-loop systems become essential. They combine the efficiency of AI with human judgment, which is especially important in high-stakes scenarios like biometric checks, edge-case transactions, and content disputes.

These systems also help mitigate algorithmic bias. AI trained on historical data is never neutral. 

It absorbs the patterns, flaws, and blind spots of the past, often replicating and amplifying societal biases. Without appropriate safeguards, such as humans-in-the-loop processes, biases can re-emerge in ways that are very difficult to detect and rectify.

Diversity within development teams is crucial in addressing this. Teams with varied backgrounds and perspectives are more likely to identify blind spots, challenge assumptions, and design systems that are inclusive from the outset. 

Tools such as active learning models and explainable AI frameworks provide further transparency, helping organisations understand why decisions are made and how to intervene effectively.

Equally important is how automation is deployed throughout the customer journey. Systems that misinterpret intent or enforce rigid protocols without room for discretion can frustrate users and damage relationships. 

Human oversight at critical junctures restores empathy and flexibility. When an AI tool misses, customers expect immediate human intervention and not prolonged escalations or impersonal responses. Trust can be gained through preventive design, but can also be reinforced through timely and thoughtful corrections.

Trust as a strategic advantage
Too many businesses still treat trust and safety as a backend function, bolted on after a product is designed. But trust is not a feature, it is a foundation that must be designed from the ground up, embedded at every point along the customer journey. It must also be reinforced by every department across an organisation, including product, marketing, and compliance teams.

A mature trust strategy means asking uncomfortable questions early-on in your process: Are we using a safety-by-design approach? Are our systems explainable? Where does responsibility lie when AI gets it wrong? The businesses that ask these questions now and on a continuous basis will be the ones that are able to better mitigate crises later.

Digital trust is the outcome of consistent alignment across people, purpose, and technology. For businesses, this means embedding trust into every layer of their operations and treating it as a fundamental to transparency, accountability, and user-centric design.