AI regulation challenges7 As AI regulation continues to evolve, there’s a need to balance the pace of technology innovation against many other factors. Progress has been made in AI policy and regulation; however, the topic remains a priority for business leaders. Countries and regions are working to establish a standard set of policies, with various groups beginning to define and publish frameworks. Examples include the Monetary Authority of Singapore (MAS) publishing a set of AI methodologies,22 the European Union’s AI Act,23 and the US President Biden’s AI Executive Order.24 Concerns persist in financial services, however, about the inconsistent approaches of regulators around the world, particularly the EU’s more comprehensive AI Act compared to President Biden’s AI Executive Order in the US. Despite ongoing efforts there are three issues that have yet to be resolved regarding current AI policy. These continue to cause uncertainty across financial services. 1. The pace of innovation: The speed at which AI is evolving and being implemented presents a challenge for businesses and regulatory bodies that lack the resources to keep abreast of the implications and emerging threats. 2. What to regulate: A related challenge is caused by the proliferation of AI models and capabilities; identifying which should be scrutinized and regulated is a mammoth task. As one size seldom fits all models, policy-makers will need to focus on the relevant use cases (e.g. objectivities, data models, responses) in order to implement successful regulatory measures.25 3. Who should regulate: The various parties have yet to agree on the appropriate level of oversight for AI policy – industry, state, federal or regional – as well as how variations of policy should be managed, especially globally.26 Though these questions remain and are presently impacting the pace of new AI regulation, some industry leaders and policy-makers have made the argument that financial services, due to its existing regulations, is better positioned than others when it comes to AI technology. A survey of 49 constituents from the Organisation for Economic Co-operation and Development (OECD), made up of various global members with experience in policy, economics and industry, highlights to some extent that current regulation is appropriate for governing the majority of AI uses. Examples of this can be found in areas such as consumer and investor protection, model risk management, third- party risk management, disclosure requirements, IT governance, cybersecurity, and fairness (see Figure 5 for additional examples).27 Artificial Intelligence in Financial Services 18 The regulation of AI will evolve as business leaders, AI researchers, policy-makers and others continue to grapple with these issues and work towards resolving gaps in current policy. 18

Artificial Intelligence in Financial Services 2025