Artificial Intelligence is redefining all sectors, and the financial sector is no exception. In this industry, the speed of technological progress is particularly evident: a study conducted by the Massachusetts Institute of Technology in collaboration with EY showed that 70% of banks are already using AI agents at some point in their adoption, especially in fraud detection and security.
In this context, LA NACION held a one-on-one meeting with Richard Harmon, Vice President and Global Head of Financial Services at Red Hat (a company that provides flexible open source solutions, including Linux, AI, virtualization and automation technologies), who visited Argentina this week. His career journey across places like Citibank, JP Morgan, BlackRock, and the New York Fed has given him a global perspective on what this industry needs to realize the full potential offered by new technologies.
-Are organizations applying AI effectively, or are they simply joining the trend?
Generative AI is being adopted very actively in various fields. now, From the perspective of AI agents (Systems that, unlike A bot Traditionally, they do not have a pre-defined written answer, but rather work with a knowledge base that enables them to perform more complex tasks and even perform actions independently. Organizations are still in a very early adoption stage. But agent technology is evolving so rapidly that it needs to be coordinated as a new workforce.
-How is artificial intelligence applied in the world of finance?
– Risky capital markets must be managed; To achieve this, it is necessary to predict what the price will be in different types of markets. There are many mathematicians and physicists studying this topic, to determine not only future value, but also sensitivity and volatility, among other factors.
A group of people set out to do just that, but using artificial intelligence agents. What was the result? They were able to come up with a formula that sets the price at the same level as the mathematicians who worked on it for a decade and not only that: they did it in a few hours.
Then, by adopting it into the workflow, time is also maximized. I know of another case where a company, including corporate clients, had to implement various tasks that took a total of five to ten days; Since they implemented automation, they can do it in two or three hours.
-How does artificial intelligence affect financial security?
-These large language models are incredibly powerful at pattern recognition. In the case of financial crimes, they can identify anomalies which the bank then investigates to see if they are money laundering or fraud.
In addition to detecting abnormal signs that humans can’t find, they generate a suspicious activity report in minutes, something that would take a person days. This saves time for the researcher, who can focus on other tasks, because he has all the necessary information, and in this way his productivity increases by 50 times.
– How will jobs in this industry change? Will they disappear?
– I see them as agents that help humans make decisions, to take into account a wider range of options. I don’t think humans will ever disappear with the integration of these technologies, but I do think they will be able to make more powerful decisions.
– What industries adopted artificial intelligence first, and is it true that it is being applied above all in the world of finance and technology?
– Financial services, if they are not at the top of the list, are placed in second place. Some areas of science, such as medicine or pharmaceuticals, are also subject to a lot of experimentation.
But there are also initiatives that seek to take care of the environment, for example. Among them, one of the cases we are working on at Red Hat is the Vital Oceans solution, which supports the protection of five marine areas in Baja California, Mexico.
– How do you see the adoption of artificial intelligence in Argentina?
– Latin America and Argentina in particular, are at the forefront of some developments; It is an area that has grown much faster than others in real-time operations (in terms of payments, transactions, etc.), which are carried out more quickly than other places. They can perform tasks from their mobile phones that are not available in the US or Europe.
A few days ago, executives from Banco Galicia, our client, were in London and told us about their future projects and I must admit they are very sophisticated. Not only that, I remember that they won an award four years ago: they used AI to process corporate requests, so when they arrived, the systems extracted, analyzed, processed all the information and performed all the checks according to the data sources and requirements you asked for. If everything is fine, the client is integrated immediately. In this way, they were able to speed up a process that took a month or two to do in less than a day.
-I worked for a few years at the Federal Reserve Bank of New York. Based on this experience, what type of regulation do you think would be good to ensure the success of AI innovation?
-On the one hand, before building something with AI, it is important to think about the end result and develop it with some safeguards in mind: transparency, trust and ethics. This way, you are more likely to have built a properly robust system.
On the other hand, in the regulatory framework, we have the EU AI law, which came into effect in mid-January. It contains specific guidelines about what can and cannot be done using artificial intelligence. I think they’ve done a reasonably good job of balancing innovation and security. They have the most complete regulations of any region, but without being prescriptively detailed.