We conclude our interview with Joris Krijger, Tech Ethicist, AI & Ethics Officer, de Volksbank, in which he asks some hard hitting questions for us all to reflect on.

In case you missed it, you can catch up on part 1 of the interview here.

I spoke to a CEO of a private institution recently, who told me that banks often forget that they're actually semi-public institutions because of their systemic importance in society. How do you look at that point of view?

I fully agree. I think banks, in general, have a role in society that is so indispensable that you cannot just consider them to be another private party selling you something. It is a fundamental good, and access to the banking system is a prerequisite to being part of society. In that sense, banks are semi-public.

And I also think the investments banks make, to a large extent, determine which sectors will have funding, and which new startups will flourish. In all of that, they should at least have the public interest or the common good in mind, because of their role, importance and power in society.

To what extent does banking also contribute to exclusion? What you see today is that people who are digitally included are also often financially in there. If you don't have access to banking, you'll actually no longer have access to society. Of course, as a bank, you have your own target markets, and you have your own customer segments that you want to serve at their best. And that often requires very specific services and products. And that's, I think, another critical dilemma, but more from a societal point of view rather than a banking point of view. Because we cannot expect every bank to serve every client, which we often do from retail banks in a way.

As I said, I think we have to re-evaluate the measurement of success and the definition of a successful bank. Right now, it's expressed financially, and it's expressed in the profits that you that you make, the return on investments that you have. In my opinion, we should really broaden that measurement of success to include societal and environmental elements.

They may be harder to measure and harder to quantify. Right now, banks are punished for serving high-risk clients because that will not reflect well on their financial performance. But perhaps you are doing a good job in terms of social performance.

As we don't have any measurements for that now, or it's not agreed as something that determines the value of an institution, I think that's where the rules of the game conflict with the common good.

I would like to see more of these measures as part of what defines success, to make sure the rules of the game will shift or fit better with what is in the public interest as well.

Your role within De Volksbank is AI and Ethics Officer. How do you see the role of AI evolve in banking in the coming years?

AI will have a tremendous impact on the financial industry. We had a couple of breakthroughs with the technology and banks have always been at the forefront of innovation and technological innovation. With AI, you see the same. And I think rightfully so because AI can, both on the front end as well on the back end, improve and automate a lot of processes. It will make banks more efficient; it will make them more customer-oriented, with faster throughput times, better estimations, better stress testing, you name it.

There are a lot of things AI can contribute to. So, in the upcoming five years, AI will really transform the way banks work. I think a lot of organisations might not be sufficiently aware of that. But it will also have an impact on the workforce and the skills required to be a successful bank in the future. Because the traditional way of doing banking will change as it did when the Internet became available. AI is considered a system technology, like the internet, like electricity, not just a new gadget, but requires a new way of thinking and a new way of working. And that's something we must get ready for.

Do you think that with AI the industry will be able to create more empathy again, in our digital channels?

That's a very interesting question. When I just started De Volksbank, five years ago, I met with a director from Pega, who was building these AI systems for customer channels. And the slogan of De Volksbank was "Bankieren met een menselijke maat", banking with the human touch.

I was very much convinced that AI would go against the human aspect. This director from Pega told me with AI, we can give an institution and organisation a memory, in the sense that they will remember who you are, what your preferences are, and what your situation is. You might argue whether this makes a bank more humane, but for sure, it becomes more personal through AI.

But AI also has its limits. It works with averages; it works with calculations, and large amounts of numbers. It's not made for exceptional situations or things that deviate too much from the mean. And in that sense, it becomes less human, I think. And that's where we will still need real humans to consider these cases that do not fit the traditional model. So, it will help to make the bank more personal. But it can also, as an effect, exclude people who don't fit the usual profiles.

I do see an opportunity to make digital banking more human, or at least create a better emotional connection again, between banks and their consumers or customers at large. AI will know which customers like proactive services and which prefer no proactive interventions. With AI, there's a potential to create that hyper personalised experience that we're talking about for 10 years now, one that knows who to offer that service and who not.

I think potentially, that could be the case. But there's this question, and I haven't seen it solved yet: “the uncanny valley”. To what extent do people like a bank that knows them very well, and knows them on an emotional level? Because, that's why it's called the uncanny valley, it can become creepy at some point. And balancing the opportunities you have with AI, with the expectations and what customer actually wants, or desires, is very difficult.

Even when it comes to people who are running into financial trouble, for example, we might assume that they will be very glad that we signal it very early on, and we'll report it to them and advise them how we can help them with the best of intentions, but not everyone appreciates that. Sometimes it backfires.

When we look at the new ethos in banking, what should I absolutely not forget in my book describing the road to the new ethos in banking, according to you?

Probably the most important point is a review of the measurements of success and the question of social responsibility. How do we define the social responsibility of a bank? Of course, AI is playing a role in that, because it's sort of reinvigorates this question. Right now, we have this technology that is not accessible for consumers. We are using it for cross-selling but also to make inferences about them. This opens the question of what is our responsibility towards consumers? So, I think that's something you must address.

And I think when you talk about the new ethos of banking, ask yourself to what extent can we change the role of the bank within society and what is expected from these banks? How can we meet society's expectations in this regard? That goes back to one of the first points I made on legitimacy. What does it mean to be a legitimate organisation, not just in 2024, but in 2034? And I think with that horizon, we need society on board.

We will need additional control mechanisms and accountability mechanisms because that's, for me at least, a critical part of trust. We need to be held accountable for the decisions that we make. At the end of last year, a report by KPMG showed that banks and insurance companies have the least trust when it comes to the use of AI and algorithms. Customers trust banks and insurance companies even less than the government with these technologies. I believe it has a lot to do with the opacity but also with the lack of accountability mechanisms when things go wrong.

So for me that's a point to focus on. Social responsibility is not just something that you claim on paper, but you have to actually dare to bring it into practice and make sure that the public can hold you accountable.

When it comes to these accountability mechanisms, you cannot say, well, a consumer should enforce his or her rights. I think we need mechanisms where we say: “Look, if we are committed to stakeholders, then why don't we give them a say in the annual board meetings?”. Yeah, right now, it's only shareholders that determine the success of the board. So why don't we give stakeholders a say in that? It will require some guts, but I think that's an important step in changing the structure because right now, we are proclaiming to care for stakeholders, but we're afraid to give them a vote in the shareholder meetings, while that would be the most direct way for them to have an impact.

So, these are the kinds of changes that I'm thinking about. (Ed: and these are the changes that we all should be thinking about too - are you? Start a conversation with our community on our LinkedIn group here).

You can hear more from Joris and meet him in person at The Banking Scene Conference 2024 Brussels - don't suffer FOMO - get your ticket today!