Financial consumer protection and market conduct considerations of AI in finance - Remarks by Colm Kincaid, Director of Consumer Protection

The following remarks were delivered at an OECD/FSB Roundtable on Artificial Intelligence in Finance on Wednesday 22 May 2024.

I am delighted to have the opportunity to speak at today’s OECD/FSB Roundtable on Artificial Intelligence in Finance. It is so important that we have these discussions and in particular consider the opportunities and risks of digitalisation and Artificial Intelligence (AI) from the consumer perspective, and what this may in turn mean for financial consumer protection and market conduct supervision.

In its 2024 Regulatory and Supervisory Outlook Report  the Central Bank of Ireland ranked AI amongst the technology with the greatest transformational potential. Over 2024/2025, we will be undertaking work to develop our supervisory expectations of regulated entities related to the use of AI in financial services. We are also looking to how AI can assist us in our work as supervisors.

We know from previous digital developments that our approach will have to be one that iterates and is progressed in close engagement with industry, consumers and technology developers. We also know that AI will affect not just financial services and the firms that provide them, but also consumer behaviour and expectations. So our approach will need to be multi-faceted. This is all the more so as societal views of technology’s proper role evolve in an increasingly ‘phydigital’ world. In addition to digital platforms, we are beginning to see the evolution not only of the algorithmic firm but also of the algorithmic consumer.

We also know from previous digital developments that the standards we set at the outset are critical to ensuring the right outcomes over the longer term. At the Central Bank of Ireland, we welcome therefore the EU proposal of an AI Act. It builds on work by the EU High Level Group on AI and indeed the OECD’s own policy work on this.  Implementing this Act will be an important task. It will bring a baseline structure and clarity to many key elements of the public discussion that has been taking place on trustworthy and ethical use of AI. We are also contributing to the setting of standards in this context through the ongoing review of our Consumer Protection Code, where we are consulting on new requirements for the provision of digital financial services.

We also, of course, have to reinforce in an AI context the standards set in financial services regulation more generally. These standards are based on lessons learned from past failures - lessons we should not forget.

So where are we focused in our thinking as supervisors of a financial service firm using AI?

Where a firm is considering using AI, supervisors will want to understand what business challenge is being addressed and why AI is an appropriate response. This means ensuring firms understand the technology they are using. It also means ensuring this technology is not applied in a way that would seek to exploit the behaviours or habits of customers where it has the potential to cause customer detriment. Data and/or algorithmic bias may lead to discriminatory customer treatment that may be difficult to identify and monitor. As supervisors, we will want to see that firms identify and prepare for any new source of risk to their operational resilience arising from the use of AI. Finally, we will look to see that there is clarity of accountability for decisions made on its use.

We also have to consider how AI informs our approach to financial education and our expectations of what financial education needs to deliver for consumers in this context. We welcome the steps being taken by the Irish Government on its financial literacy strategy and the ongoing work at international level  by the OECD and the International Network on Financial Education (INFE).   We also welcome the broad assessment of skills and capability needed by users of AI including in financial services, as well as other settings like the public sector, and the broad framing of this in the Irish Government’s National AI Strategy.

In a world where consumers have so many new ways of interfacing with financial services and advice, we need to equip consumers not just to understand the basics of financial services but also the risks to which they are exposed in digital environments. Most obviously this includes the heightened risk of fraud, but the risks in which consumers need to be educated regarding AI go much wider.

We also need to monitor closely whether the use of AI leads (inadvertently or by design) to the exclusion of people from financial services they would otherwise have had access to.

Fundamentally, financial services are based on trust and we need to ensure that the use of AI does not damage that trust. We also need to recognise that AI has great potential to enhance trust through the provision of better financial services. A key part of this will be ensuring firms understand and take ownership of design biases in AI.

And of course, public authorities will be both users and deploy AI. We need to practice what we preach and ensure our use is trustworthy.

Artificial intelligence has great potential to improve the service consumers of financial services receive.  The work of public authorities and commercial enterprises together should be to ensure that AI is used to bring those improvements about and that AI is not used to consumers’ detriment. Financial regulators have a specific and important role to play in this dynamic.  By standing our ground on the standards we require when it comes to AI, while being open minded about AI’s potential, we can provide the platform for AI to flourish in a way that provides real and sustainable long term benefits for consumers of financial services, today and in the future.

Thank you again for inviting me to speak today and I look forward to our discussion.

Thanks to Trevor Fitzpatrick, Deirdre Mullally, Johanna Zelenak and Nicola Faulkner for their comments and support in preparing these remarks.