ISFB Insight
When Machines Enter Language
June 3, 2026
Artificial intelligence is becoming part of the workplace lexicon. In the banking sector, it will likely shift tasks rather than eliminate entire job roles. It will make the fundamentals more apparent, forcing organizations to provide extensive training to their teams and requiring each employee to further develop their technical skills.
We tend to view artificial intelligence as part of the long line of automation advancements that have transformed work, and there is something reassuring about that comparison. As early as the last century, people were already seeking to streamline physical tasks; Frank and Lilian Gilbreth’s work on the analysis of movements aimed to eliminate unnecessary steps to increase efficiency. Machines took over the physical tasks, while individuals retained the ability to speak, reason, and make decisions.
Today, AI no longer just performs actions; it is becoming part of language. It writes, rephrases, summarizes, compares, and prepares decisions. Yet language is not just another tool: it is the very medium through which we think. When a machine becomes part of language, it simultaneously becomes part of the way an organization reasons, makes decisions, and tells its story. In short, it becomes part of its organizational culture.
The task rather than the profession
People often wonder which jobs might disappear—and how soon. I’m skeptical of figures that claim to predict the future. Here, it’s important to distinguish between the short term—where we can make projections with an acceptable margin of error—and the long term, which falls under the realm of foresight—and sometimes even a form of futurology, whether acknowledged or not. Forecasting is useful, but it must be taken for what it is: a body of scenarios based largely on economic data—which often lags behind what it purports to describe—and collective imaginations. Nevertheless, it has the merit of encouraging organizations to invest time in strategic thinking.
Finance is not simply a list of jobs that can be eliminated or retained based on the power of the tools. AI is already being integrated into certain processes—such as compliance, portfolio selection, fraud detection, and cybersecurity. It assists, filters, accelerates, and can improve the quality of work. However, it does not resolve the issue of assessing a situation. In other words, the debate is not about humans versus machines, but about tasks supported by technology. In a bank, opening an account, recommending an asset allocation, assessing a risk, or taking responsibility for a decision remains a human endeavor, because it involves accountability, rules, and a relationship that is both personal and institutional.
The Paradox of Fundamentals
AI should not become the one-size-fits-all solution for every current challenge. Not all aspects of banking fall under the purview of AI. The majority relate to financial literacy, organization, management, training, law, or customer relations.
Tomorrow, however, in knowledge-based professions, simply producing content will no longer be enough. Machines will help us produce faster, more cleanly, and more accurately. Value will shift toward understanding, putting things into perspective, and interpretation. To learn about a topic, I can read a brief article in a free daily newspaper or look up the seminal paper, followed by a few meta-analyses. I won’t gain the same understanding. It will all depend on my sources, of course, but also on my ability to connect what I read to other knowledge, to recognize its limitations, to identify biases, and to avoid confusing a clear answer with a sound one.
AI will amplify the capabilities of those who already havea strong relationship with knowledge, and will tend to undermine those whose relationship with knowledge is less developed, in an environment where the speed of information dissemination very easily masks superficiality. AI will not create skill gaps, but it will sometimes make them more pronounced. Training can help mitigate these differences, provided it is not aimed solely at the most curious individuals.
Toward a Skills Transformation
The question that organizations are openly asking themselves remains: Should we move quickly, or proceed cautiously? If we move too slowly, we risk falling behind; if we move too quickly, we’ll make the mistakes of youth. What seems certain to me, however, is that banks must provide extensive training to their teams without further delay. Not just to write better prompts, but to understand how AI could transform banking: customer relationships, the quality of advice, operational processes, compliance, data security, legal liability, and even professional judgment. Employees, too, have every reason to take the initiative, because AI will change their jobs, whether they chose it or not.
That is precisely the purpose of the summer training program offered to banking professionals, and which can also be implemented internally at ISFB member institutions to incorporate the internal cultural dimension.
Training now begins with choosing a direction. A bank must be able to specify what it expects from these tools, what it rejects, what it is testing, and what it keeps under control. Otherwise, everyone risks tinkering on their own, sometimes using the organization’s tools, sometimes using their own.
Proper use will undoubtedly come through successive trials in controlled environments, with the freedom to test without putting the entire institution at risk. Management must provide clear guidance on where AI belongs, where it does not, and what decisions should continue to be made by human judgment.
The Price of Relief
One final point strikes me as important, because it touches on what gives a piece of work its value. AI doesn’t just produce content; it standardizes it. We’ve all come to recognize these polished texts that all have the same flavor, the same words, the same syntactic structures. It provides access to a respectable average quality, but at the cost of flattening styles and reasoning. Yet what no longer has depth or distinction eventually loses its value.
AI will not replace humans. It willsimply become part of nearly every tool used in intellectual work. And, like any tool, it says nothing on its own about the quality of the work produced. A hammer can be used to build straight walls or to damage them. AI raises the same question: used without discernment, it will make us faster and likely more superficial; used with discernment, it will help us better compare, verify, formulate, and decide. This is also what makes banking education even more necessary than ever before: the easier access to information becomes, the more we need to know how to understand it, contextualize it, and assess its value.
This may be the least spectacular task ahead of us once we finally stop talking about AI and refocus on the fundamentals of our professions. It will be the most demanding but also the most rewarding: using our tools without ceasing to think, knowing when to slow down, staying attentive to our relationships, and not falling into the trap of taking the easy way out.
© Institut Supérieur de Formation Bancaire (ISFB). All rights reserved.
Analyses and content published by the ISFB may be cited or reproduced in part, provided the source is clearly acknowledged. Any full or substantial reproduction of this article in another medium or format is subject to prior written authorization from the ISFB. To facilitate reading and without any intention to discriminate, the masculine gender is generally used, in accordance with the grammatical rule that allows it to be used as a neutral term to refer to a group of people comprising both men and women. This publication is intended for ISFB members and their employees in Switzerland, as well as anyone interested in finance in Switzerland. It is not intended to be read or distributed in any jurisdiction where its distribution would be prohibited.
