After Singapore Fintech Festival 2025:Why AI Assurance Is Becoming Finance’s Most Urgent Frontier

By Hongzhe Liu, Co Founder, Ailantis Pte. Ltd.

The Singapore FinTech Festival this year carried a very different energy. Artificial intelligence was no longer an experimental add on or a distant idea. It had become the centre of gravity for almost every conversation. What once focused on digital payments or financial infrastructure now revolved around responsible design, clear guardrails, trusted data foundations, and the larger question that every financial institution is beginning to ask. How do we trust artificial intelligence enough to scale it?

Across panel discussions, private roundtables, and the exhibition halls, one theme consistently stood out. Artificial intelligence has accelerated, but risk management has not evolved at the same speed. In his keynote remarks, Deputy Prime Minister Gan Kim Yong emphasised that artificial intelligence offers major gains, but without careful planning and oversight, it may create more harm than good. For Singapore, a country known for its careful regulatory leadership, the message was unmistakable. Innovation must progress together with strong governance.

The financial sector illustrates this reality vividly. Banks, insurers, asset managers, and service providers are already using artificial intelligence in many critical areas. These include fraud monitoring, credit evaluation, customer service automation, and compliance checks. What began years ago as simple rule-based tools has now developed into advanced learning systems that create content, interpret images, and generate real-time decisions. DBS shared during the festival that it now operates more than one thousand separate artificial intelligence models. These systems depend on governed data lakes, clearly defined protocols, and a culture that treats data responsibility as a priority. According to the bank’s leadership, a major part of this journey required a complete shift in mindset. Teams had to learn again, experiment openly, and develop shared responsibility across the organisation.

Yet as artificial intelligence becomes more embedded, its risks grow. Unlike traditional software, an artificial intelligence system can behave in unpredictable ways. It may drift over time, produce inaccurate content, or demonstrate bias that was not present previously. These issues can lead to serious consequences, including regulatory breaches, financial losses, reputational problems, or legal claims. Regulators in many regions, including Singapore, the European Union, and the United States are raising expectations around responsible use. However, a large number of enterprises still lack structured methods to evaluate whether their artificial intelligence systems are safe or compliant. 

This brings insurers to the centre of the conversation. Financial risk has always been their domain, but artificial intelligence presents a completely new type of exposure.

Traditional cyber insurance focuses on attacks or breaches. Traditional professional liability insurance focuses on human errors. Neither of these categories was designed to address the failure of a learning system that may produce incorrect results even when there is no attack and no human fault. For insurers, artificial intelligence represents both a major opportunity and a complex challenge. They need a way to examine, score, and understand these models before they can provide accurate coverage.

This is the role of artificial intelligence assurance. Throughout the festival, regulators, banks, technology companies, and professional service firms all pointed to assurance as the critical missing piece. If organisations use artificial intelligence faster than they can govern it, then assurance becomes the central mechanism that allows boards, auditors, and insurers to trust these systems.

Among the companies contributing to this emerging field is Ailantis, a Singapore based platform focused on artificial intelligence assurance for the insurance sector. Rather than presenting itself simply as a governance tool, Ailantis focuses on the specific question insurers need answered. How do we quantify the risk of an artificial intelligence system in a structured way that supports underwriting decisions? The platform evaluates generative models, computer vision systems, audio models, and other learning algorithms through controlled testing, simulated stress events, and performance scoring. The output is a clear risk profile that expresses reliability, exposure severity, and model behaviour under challenging conditions. 

The message from the festival is aligned with this approach. Artificial intelligence governance cannot depend on intuition or informal review. It requires evidence that can be measured and checked. 

It requires systematic testing that shows how a model behaves when data changes or when it faces difficult scenarios. 

For insurers, this kind of transparency creates confidence where there was previously uncertainty. The most striking insight from the festival, however, was not only that assurance prevents harm. It can also accelerate innovation. When risks are known and described clearly, financial institutions are able to adopt artificial intelligence with greater confidence. Regulators gain assurance that safety measures are in place. Insurers can create new categories of protection including artificial intelligence liability and artificial intelligence warranty coverage. In many discussions, industry leaders described assurance as a new competitive advantage.

Singapore is positioning itself at the centre of this movement. The new partnership between the Monetary Authority of Singapore and the United Kingdom Financial Conduct Authority reflects a growing global commitment to responsible artificial intelligence. Together, they plan shared testing, regulatory knowledge exchange, and industry guidance. SFF 2025 demonstrated that the future of finance and the future of artificial intelligence are now inseparable. With strong assurance, artificial intelligence becomes not a source of uncertainty but a trusted asset that the financial system can scale and insure.