AI & Banking Webinar Series

In September and October 2025, the Centre for Banking and Finance Law (CBFL) at the National University of Singapore, Faculty of Law co-hosted the “AI & Banking Webinar Series”, together with the University of Reading and the University of Johannesburg. The six-part webinar series brought together leading legal experts from diverse jurisdictions and specialties to explore the promise and perils of increasing adoption of AI in banking operations.

Session 1 (9 September 2025): AI in Banking and Finance – an Overviews

Moderator: Dr Sandra Booysen, National University of Singapore

Panellists: Prof Teresa Rodriguez de las Heras Ballell, Universidad Carlos III de Madrid; Ms Marzia Briel, University of Reading; Assoc Prof Hannah Yee-Fen Lim, Nanyang Technological University, Singapore; Assoc Prof Lin Lin, National University of Singapore

Associate Professor Sandra Booysen, the director of CBFL, opened the webinar series by noting the emergence of AI and the fact it presents opportunities and poses risks, including in banking and finance. The first panellist, Associate Professor Lin Lin, gave a short presentation about the use of AI by Chinese banks, noting its potential to address structural inefficiencies. A fruitful discussion followed among the panellists. Professor Teresa Rodriguez de las Heras Ballell and Associate Professor Hannah Yee-Fen Lim engaged in a lively discussion about whether a distinction could, or should, be made between AI as a product and AI as a service. Ms Marzia Briel noted AI’s applicability to climate and biodiversity matters but expressed scepticism about whether ethics-based approaches were adequate to manage AI risks. The discussion finally turned to the EU’s AI Act and whether it helps or hinders innovation.

Session 2 (16 September 2025): AI and Money Laundering

Moderator: Dr Cayle Lupton, University of Johannesburg

Panellists: Mr Eric Chan, Singapore Deposit Insurance Corporation; Prof Nicholas Ryder, Cardiff University; Dr Loice Takaedza, University of Johannesburg

Dr Cayle Lupton began the webinar by noting that anti-money laundering (AML) is an area where AI could make a major difference, given the persistent problem of false positives. Professor Nicholas Ryder first presented an overview of current approaches to AML. He noted that AI could enhance transaction monitoring provided human investigators remained in control. Next, Mr Eric Chan focused on concerns related to AI, including the “black box” problem of inexplicable outputs. Mr Chan also commented on “AI versus AI” scenarios, where malicious actors also use AI to perpetrate crime. Next, Dr Loice Takaedza introduced South Africa’s AML framework. Dr Takaedza noted that though South African banks currently use AI to comply with AML obligations, its use is neither addressed by AML rules nor any general AI laws. The webinar concluded with a Q&A session in which panellists discussed “defensive” AML reporting and the situation of small financial institutions.

Session 3 (24 September 2025): AI and Consumer/SME Credit

Moderator: Dr Monica Vessio, University of Reading

Panellists: Prof Jodi Gardner, University of Auckland; Ms Nonkululeko Sefolo, University of Johannesburg & TransUnion Africa; Assoc Prof Jon Mark Truby, National University of Singapore

Professor Jodi Gardner began this webinar with a presentation on tort law’s potential to address harms arising from algorithmic decision-making. Algorithms are especially pervasive in consumer credit, making systems of responsibility important. Next, Associate Professor Jon Mark Truby commented on the use of AI in loan transactions: though accountability, transparency and data protection are challenges, Associate Professor Truby suggested that the use of AI could expand access to finance, improve efficiency and thus support sustainable development. Ms Nonkululeko Sefolo then commented on AI in credit scoring, drawing on the example of TransUnion Africa, a credit bureau. Ms Sefolo described the use of AI for testing and simulation, and the potential for alternative types of data (e.g. from telecommunication service providers) to be used to assess creditworthiness. After a short Q&A session, Dr Monica Vessio concluded by observing that AI in banking is a normative development, not just a technological one.

Session 4 (1 October 2025): AI and Financial Advice/Analysis

Moderator: Dr Monica Vessio, University of Reading

Panellists: Dr Alessio Azzutti, University of Glasgow; Dr Kolja Gauer, Otto-von-Guericke University; Mr Eduard Buzila, Otto-von-Guericke University

This webinar began with a presentation by Dr Alessio Azzutti on the use of AI in capital markets, including under the EU’s MiFID II regime. Since MiFID II is “technology-neutral”, this could enable the use of AI. Dr Azzutti also suggested that AI could be applied as “RegTech” (a compliance aid for firms) or “SupTech” (an enforcement aid for regulators). Next, Dr Kolja Gauer and Mr Eduard Buzila presented an ongoing research project in which they investigate the use of AI by lawyers in drafting prospectus disclosures. Through a multi-step, mixed methods investigation, their project assesses whether generative AI can be used in prospectus drafting, and how it compares to humans in terms of quality. The panellists then addressed questions on the regulation of AI risks, risks to consumers, and the impact of value judgments on research design.

Session 5 (7 October 2025): AI and Trade Finance

Moderator: Dr Cayle Lupton, University of Johannesburg

Panellists: Dr Tinaye Chivizhe, University of Johannesburg; Emeritus Prof Charl Hugo, University of Johannesburg; Assoc Prof Dora Neo, National University of Singapore

Emeritus Professor Charl Hugo began this webinar with a presentation on the doctrine of strict compliance in documentary credit law, a key part of trade finance. It was noted that the doctrine is not absolute but leaves little room for manoeuvre. Dr Cayle Lupton added that AI could be used, in this context, either for sanctions-related matters or to check for documentary compliance – although the latter use of AI is still in its infancy.  Next, Dr Tinaye Chivizhe commented on AI and document examination in the African Continental Free Trade Area (AfCFTA). Dr Chivizhe noted a lack of rules governing AI under either the AfCFTA or commercial standards like the eUCP. Associate Professor Dora Neo then returned to the potential use of AI in documentary compliance, suggesting anecdotally that industry confidence is low, but that the UNCITRAL Model Law on Electronic Transferable Records could offer a way forward. The session concluded with a brief round of questions.

Session 6 (14 October 2025): AI and Payment Fraud

Moderator: Dr Sandra Booysen, National University of Singapore

Panellists: Dr Ilias Kapsis, City St George’s, University of London; Dr Sharon Munedzi, University of Johannesburg; Assoc Prof Andrew Mazen Dahdal, Qatar University

Associate Professor Sandra Booysen opened the final webinar by noting the significance of payment fraud and the harms caused both downstream and upstream. Associate Professor Andrew Mazen Dahdal presented some of the features of UK law and how they respond to payment fraud, suggesting that AI could be used to assess the vulnerability of bank customers. Dr Sharon Munedzi introduced the topic in a South African context, suggesting that there is currently a bits-and-pieces legal framework and that the gap between law and technology must be bridged. Dr Ilias Kapsis then looked at fraud from an EU angle: AI poses major risks by making it easier to perpetrate fraud, such as by defeating voice recognition mechanisms. Dr Kapsis suggested AI could also be used preventively, especially to flag unusual behaviour. In the final Q&A session of the webinar series, the panellists discussed legal obstacles to the adoption of AI, whether South Africa needs an AI law of its own, and where AI could best be deployed along the fraud prevention chain.