BIS’s Hernández de Cos on AI, non-bank risk and tokenisation

When you were speaking at the annual report press conference in June, you mentioned the BIS 100 strategy was a response to the complexity of the global environment. Can you say a bit more about what you mean and what the strategy is trying to respond to?

When I joined the BIS more than a year ago, the previous strategy cycle, led by Agustín [Carstens], had just been successfully concluded. From day one, I began working with my management team and BIS staff on a new strategy. We call it BIS 100, as it will conclude in 2030, coinciding with the centenary of the BIS.

The first step involved analysing the initial position of the institution and identifying the key megatrends relevant to the BIS. One important aspect is the increasing complexity of the economic and financial environment. This is marked by uncomfortably high levels of public debt, the growing role of non-bank financial intermediation, and more frequent and severe supply shocks. Another key megatrend is technology, particularly the rapid pace of development. This includes major advances in artificial intelligence and financial innovation related to tokenisation and new forms of digital money. Also, like any international institution, the BIS has to navigate ongoing geopolitical tensions.

BIS 100 has been designed to address the complexity of this global environment and further strengthen the way we support our member central banks in pursuing their price and financial stability mandates. BIS 100 aims to solidify our strengths in four core activities: providing sound and timely economic and policy insights; fostering dialogue among central banks through our convening power; examining and experimenting with new technological developments; and providing banking services to the central bank community.

Pablo Hernández de Cos has been general manager of the BIS since July 2025. He was governor of the Bank of Spain and a member of the European Central Bank’s governing council from 2018 to 2024. He chaired the Basel Committee on Banking Supervision from 2019 to 2024.

Hernández de Cos began his career at the Bank of Spain in 1997, where he held roles including director-general for economics, statistics and research. He was a professor of economics at IESE Business School, adjunct professor at University Carlos III of Madrid and fellow of the think-tanks Bruegel and Peterson Institute for International Economics.

To pick one element of the strategy to begin with, what do you see as some of the most important questions for the monetary and economic department right now? Is there a shift in research emphasis?

We have identified four main areas of research on which the institution will focus.

The first is technology and its implications for the future of money. This includes developments in tokenisation, ranging from stablecoins to tokenised deposits.

The second is global finance, particularly the growing role of non-bank financial intermediaries in intermediating increasingly large amounts of public debt, as well as their interconnections with the traditional banking system.

The third area is structural change. Here, we plan to focus on the consequences of geopolitical fragmentation for the global economy, the effects of demographics, or population ageing, and the impact of artificial intelligence on productivity, economic growth and the labour market.

The fourth area is financial regulation. An important goal is to assess more fully the effectiveness and potential side effects of different regulatory measures. This type of analysis is essential to inform the ongoing debate on modernising regulation.

I think some people may be concerned that the push to ‘modernise’ regulation could be used as a cover for deregulation. Are you confident that it will be possible to find efficiencies without jurisdictions actually undermining the core of the regulatory framework?

Let me start by saying that, as we approach the 20th anniversary of the global financial crisis, I think it is important to assess the effects of the regulatory measures introduced since then to prevent such a severe crisis from reoccurring.

I also think most people would agree these measures have been very successful in making the banking sector more resilient. Banks have been able to withstand a series of major shocks, including the Covid-19 pandemic, Russia’s invasion of Ukraine, trade tariffs and the ongoing conflict in the Middle East. The resilience of the banking sector has supported the resilience of economic activity.

But while financial regulation has strengthened banking resilience, it has also imposed a significant administrative burden on banks and regulators. It is important to think about whether this burden can be reduced – for example, by simplifying regulatory frameworks – while preserving resilience. There are several avenues that could be explored, some of which I discussed in a speech I made in Geneva earlier this year.

When considering ways to refine regulatory frameworks, it is critical to make this a forward-looking exercise. The financial system is not static. It is highly dynamic. This means the regulatory debate should pay close attention to emerging trends, such as the growing role of the non-bank financial intermediaries I mentioned earlier, digital innovation and increasing cyber security risks.

Finally, let me stress one point that we at the BIS hold particularly dear. In today’s interconnected financial system, financial stability is a global public good. International co-operation on financial regulation is, therefore, essential.

The BIS has devoted a lot of effort to understanding the migration of risk from banks to non-banks, and the connections between these sectors. What’s your sense of how risks have evolved? On the one hand, you arguably have a more resilient banking sector thanks to Basel III, but you also have a lot of leverage in the non-bank sector and many less regulated firms operating in that space.

Starting with the positive, the banking sector has proved highly resilient in recent years, largely owing to the strength of the regulatory framework. However, the growth of non-bank financial intermediaries (NBFIs) is creating new financial stability challenges. This is particularly concerning because some NBFIs, especially highly leveraged hedge funds, play a critical role in the intermediation of public debt.

As we consider appropriate policy responses, it is important to recognise that NBFIs are often closely connected to the banking sector. This happens, for example, through the repo and FX swap markets. Therefore, effective bank regulation also helps to make NBFIs resilient. And, most importantly, it reduces the risk that disruptions in the NBFI sector could jeopardise the stability of banks.

In today’s interconnected financial system, financial stability is a global public good. International co-operation on financial regulation is, therefore, essential

Pablo Hernández de Cos, BIS

Besides preserving the resilience of the banking sector, it is also important to establish congruent, or equivalent, regulation across banks and non-banks. The guiding principle is that regulatory frameworks should apply a similar degree of stringency to all financial intermediaries that pose similar risks to financial stability, regardless of their legal form or business model.

The BIS has been warning for a long time about fiscal policy, and the situation only seems to have gotten worse. You identify in the 2026 annual economic report how a ‘fiscal-financial nexus’ is reducing central banks’ room for manoeuvre. But is there anything central banks can actually do about that, given the political backdrop?

High levels of public debt pose considerable challenges for monetary policy. As we explain in our latest Annual Economic Report, this is because they can alter the way monetary policy is transmitted to the real economy and increase the risk of abrupt market reactions.

The first and most important line of defence is for governments to understand fiscal consolidation is of the essence. Correcting fiscal imbalances is critical to ensuring that monetary policy can successfully fulfil its mandates. A second important priority is sound financial regulation across both banks and non-banks. This limits the risk that fiscal imbalances could trigger financial stability problems. Finally, central banks should be prepared to intervene if there is market dysfunction. Properly designing emergency intervention programmes is essential for such measures to succeed as well as to contain potential moral hazard effects.

I wanted to turn to some of the technology that you mentioned and firstly ask to what extent is the BIS using AI in its own work? And how are you managing AI risk?

As you can imagine, this is a highly dynamic area. Like many other institutions, we are trying to make the most of this important technology in our daily work. For example, AI tools are already providing valuable support for our research. Our Innovation Hub has also launched several projects to examine different aspects of the technology and draw lessons about how central banks can use these tools most effectively. And our banking department is exploring using such tools in market analysis.

Correcting fiscal imbalances is critical to ensuring that monetary policy can successfully fulfil its mandates

At the same time, we are keenly aware of the operational risks associated with AI, particularly given it is at an early stage of development. Like most of our member central banks, our tolerance for operational risk is very low. We are rolling out AI at a measured pace so we can minimise the associated risks.

On the broader macroeconomic impact of AI, in your recent speech in India you set out three scenarios for AI’s impact: a permanent upward shift in the growth trend; a demand bottleneck; and explosive growth. Do you have a sense for which is more plausible?

Predicting the medium- and long-term impact of AI on economic growth is obviously very difficult. It will depend largely on how the technology develops and the policy responses to it.

The explosive-growth scenario is the most optimistic. It assumes AI will improve autonomously and generate big gains in productivity. The upward-shift scenario also involves productivity gains, but on a much more modest scale. This scenario is broadly consistent with central estimates in the literature. They suggest that AI could increase productivity growth by approximately 0.5 percentage points per year. This would still be a substantial gain if you compound that over time. It also appears plausible when you think about how much AI has been found to improve productivity in specific tasks. But there is still a lot of uncertainty about how much these gains can translate into increases in overall productivity.

The third scenario also assumes AI can generate significant productivity gains. However, it shows that if AI increasingly shifts income away from workers by substituting for labour, this could weaken aggregate demand and ultimately constrain economic growth. To offset this risk, policy-makers could consider ways to redistribute income and invest in workers’ retraining and reskilling. This would reduce the disruptive effects of labour substitution.

There has been a lot of concern among central banks about cyber risk, which to some extent is driven by AI. Do you see a role for the BIS there in terms of co-ordinating a response to cyber threats?

Indeed, we have seen an increase in concerns about cyber risks in recent months, largely because of developments in AI. I believe the BIS has an important role to play in supporting central banks in managing these risks. This is because cyber security, like financial stability, is a global public good. Given the interconnections between financial institutions and payment systems, a severe cyber incident in one jurisdiction can quickly spill over across borders.

The BIS is well positioned to help central banks discuss these risks, develop appropriate cyber security protocols, and identify areas for collaboration. The standard-setting bodies hosted by the BIS are themselves working on these important issues. And our Innovation Hub can serve as an important laboratory for examining cyber risks and potential responses.

How would you assess the impact of the Innovation Hub so far? It was obviously a very major focus for your predecessor as general manager, Agustín Carstens. It remains a part of the strategy, but it now sits alongside these other areas. How has the hub evolved?

Agustín and his team were visionary in recognising how critical innovation is for the central banking community. The Innovation Hub was created precisely to help central banks understand the potential benefits and risks associated with technological developments and to experiment with them.

As I mentioned at the beginning, we began developing our strategy by assessing our current position. We also asked ourselves if the Hub had performed as intended. In my view, the Hub has successfully focused central banks’ attention on important technological themes and helped inform key policy discussions. The Hub is still relatively new and has a lot of further potential.

A key challenge for the Hub is to keep pace with technological developments across a complex range of topics. These include AI, tokenisation, cyber security and quantum technologies. Yet this challenge is also a key reason for the Hub’s existence, as it helps the central banking community join forces to keep up with emerging trends.

Some of the most interesting projects that have resulted from the Innovation Hub are Agorá and mBridge (though the BIS has stepped away from the latter). Both are focused on wholesale settlement, but in a sense, they seem to have been affected by the geopolitical environment. It looks like you have one bloc aligned with China that is leading the work on mBridge and then a sort of Western bloc working on Project Agorá. At the moment, there seems to be no technology to bridge between the two. Is that a problem?

We have a multi-pronged approach to our projects and sometimes exploring the different technical solutions is done more efficiently at the regional level. In other aspects, it requires a more global perspective.

Policy responses that are not co-ordinated could generate a risk of fragmentation, which in turn could negatively affect central banks’ ability to deliver on their global financial stability objectives.

A key challenge for the Hub is to keep pace with technological developments across a complex range of topics. These include AI, tokenisation, cyber security and quantum technologies. Yet this challenge is also a key reason for the Hub’s existence

And looking at the technology behind Agorá and mBridge, do you have a sense for which of those projects is perhaps a more promising method?

We are neutral in terms of the technology itself. Our role is to explore and to analyse the different advantages and disadvantages of the different technologies, and draw lessons from this.

In the particular case of stablecoins, we are stressing that if stablecoins aspire to be money, there are, at least today, some serious challenges that we will need to solve. To guarantee integrity is a very important one. The lack of scalability is another, as is the lack of singleness. The policy conclusion is that the current arrangements related to stablecoins have to be reinforced to improve all these different dimensions. And we are helping achieve this objective. For example, Project Aurora by the Innovation Hub explores data-driven approaches to help combat money laundering.

We also see that the technology itself, tokenisation, could have very important advantages. And that by incorporating this technology in the current two-tier monetary system, some of these drawbacks disappear while the benefits remain. In this vein, Project Agorá has been testing the possibility of using tokenised deposits and tokenised central bank reserves at the core of the system for wholesale cross-border payments. We’ve ended phase one. And the outcome of the exploration done by the BIS, together with a group of central banks, plus the commercial banks, is rather positive. Of course, there are still many questions to be answered, and this will be the subject of future phases of the project.

Do you think it is important for central banks to move forward with work on tokenisation to ensure that you don’t end up with a fragmented system, with ‘walled gardens’ created by private sector firms that might be issuing stablecoins or tokenised deposits on their own?

Yes, and the BIS is helping the central banking community to achieve precisely this objective. In particular, we are focused on providing a common understanding on how this technology can be useful for the type of purposes that you are mentioning and on ensuring there is some degree of co-operation.

I wouldn’t take for granted this idea of fragmentation being the final result. The role of the BIS is trying to avoid such a scenario, given the potential negative implications for financial stability associated to it.

I wanted to ask about the BIS banking department, which is another focus of the new strategy. I know Agustín Carstens oversaw changes in the technology used in the banking department and new banking products. What are your plans now?

If we look back at the last 10 years, the transformation of the banking department has been very significant, including in using new technologies, developing new products, and in the end, providing services to the central banking community. Both the volume and the quality of the services we provide to the central banking community have increased. The demand for our services remains very strong.

In the new strategy, we want to focus on responding to this demand in a more tailored manner. This requires a more specific analysis of the needs of the different central banks so that we can accommodate their requests while preserving, of course, the risk management of the whole institution. At the same time, the transformation of the department has also been significant, and we have also to guarantee the operational resilience of the banking business.

One thing the Annual Report showed this year was that there was a lot of demand for BIS banking services from central banks in Asia and Latin America. Can you say anything about the particular reasons that are driving that?

If you draw a map of where the reserves of the global economy are distributed, then to a certain extent there is a correlation, a certain connection with the demand for our services.

Lastly, you are widely seen as a frontrunner for the ECB presidency. If you do get the job, what lessons would you take from your time at the BIS?

I don’t have any comments on that. As you can see, I’m very much focused on the agenda of the BIS. We’ve just designed our strategy for the next four years, and we are focused now on delivering on it.