Opening remarks at the Conference on AI and Systemic Risk Analytics

Helsinki, 6 June 2024

Good morning, ladies and gentlemen.

It is my great pleasure to welcome you all to today’s conference on AI and Systemic Risk Analytics. I am especially happy to see so many of you here in Helsinki. I would also like to warmly welcome all of you who are participating online.

The topic of this year’s conference, AI and Systemic Risk Analytics, is very timely and, of course, highly interesting. AI and data centricity have developed rapidly in recent years and the pace is picking up. Increasing use of AI also means a paradigm shift in the use of data – more data is constantly needed for AI.

The pace of this technological change challenges our knowledge – novel technologies and use of data could profoundly change the functioning and structure of economies, finance and society in general. The key challenge is to keep pace with AI and to deepen our knowledge in order to produce relevant, reliable and well-targeted analysis, policies and regulation.

---

AI has the potential to transform transmission channels, interactions and amplification mechanisms in the financial system. As the operations in financial systems become more automated and complex, there is a need to understand how the truly systemic level of risks evolves with this change. To do this we need better and more granular data with wider coverage. Data should already be seen as a strategic priority. Similarly, as the environment gets more complex, understanding the key vulnerabilities requires more and more cooperation, for example between regulators and central banks as well as with data registries.

---

Identifying and understanding systemic risks is at the very core of designing effective and purposeful policies and regulation for the financial system. Understanding the novel risks underlines the need to ensure constant knowledge building and skill-set updating for policymakers and regulators. They need to understand how the financial system vulnerabilities are evolving and transforming. To do this, they should be able to use the same technologies in AI as the market operators, and they should have the ability to challenge any inaccurate and unexplainable output generated by AI.

In terms of risk management, the growth in the use of AI has sometimes been likened to the period preceding the global financial crisis. No wonder, as there are clear similarities.

During the global financial crisis a vast number of new risk managing methods emerged. They were widely used globally. However, their vulnerability was that they failed to capture the rapid growth in interdependencies, the hidden risk concentrations and complex interconnections, as well as the true nature of the risks and risk transfers, many of which were largely misunderstood and mispriced.

What I hope we have learned from the global financial crisis and subsequent crises is that 1) in order to maintain trust we need good visibility to determine how the interconnections, linkages and risks have shifted in the financial system and who ultimately carries the risks; 2) we need to preserve accountability and increase our understanding of AI technology, its limitations and uses – we cannot have too many black boxes in the financial system if we want to maintain trust; and 3) we need to understand what will be the impact of our own actions on policies or regulation.

---

In addition to relearning lessons from the past, there are also very novel features in AI-related risks. For example, AI enables and amplifies the successful execution of scams and cyberattacks, such as in the case of advanced deepfakes and the rapid spread of disinformation. This is particularly relevant now when geopolitical instability could increase the probability of cyber incidents.

As the use of AI increases, AI literacy also needs to be improved in order to effectively address and tackle AI-related crime and misinformation. Questions related to data protection, data privacy, data management and the ethical use of AI are all important too. The significance of data security, in particular, has increased in recent years. Data breaches have become more and more common. As the amount and use of data increases, it becomes more important to have some knowledge of data security and of how to securely control your own private data.

It is essential that we continue and further develop the assessment of risks caused by AI. As the world becomes increasingly digitalized, responses become faster and impacts are felt sooner. This can be both good and bad. Faster reactions occur in the market. As we saw in the case of deposit runs, this requires rapid stabilizing actions from regulators and central banks. The challenge is to ensure that the stabilizing regulatory and policy tools are fit for purpose and ready to be used.

Finally, the use of technology and AI could cause fragmentation in the financial system and in the distribution of economic resources. It can divide industries and operators in a new way, into those which are declining and those which are growing, depending on their knowledge and ability to use the new technologies. The same can happen in other areas of society. That is why it is important to have a majority of the public on board in the technological shift and to guarantee a certain level of technological literacy for everyone.

I would like to finish with a few words on the key topics of the conference.

First, I must congratulate the organisers for compiling such a topical and wide-ranging programme for this conference. I would like to warmly welcome the accomplished keynote speakers and all the presenters to the conference.

Over the coming two days we will cover various topics related to data and AI, concerning their role in shaping the economy, finance and banking. We will take a closer look to the role of AI in the securities markets, and we will look at phenomena such as decentralized finance (DeFi). We will also learn which are the hot cybersecurity topics for the year to come. And we will hear how quantum computing could change the risk outlook. Just to mention a few of the topics.

For financial stability and macroprudential policy makers like myself, this area of research offers invaluable insights.