By Diego Patricio Cazorla, President – Instituto Argentino de Ejecutivos de Finanzas (IAEF), April 2025
Experience to date indicates that the rollout costs of AI projects, given the practice acquired by companies during the successive technological changes of recent decades, will prove to be a reasonably predictable and manageable factor.
This article aims to briefly alert CFOs and their teams to how their role in organizations will be impacted by the world of Artificial Intelligence (AI). It’s not about the technical issues; it’s about being concerned about organizational behavior with the inclusion of AI across all organizational functions, including those outside their scope.
While other colleagues focus on the technological challenges, CFOs must focus on ensuring the organization is ready to adopt it. To a greater or lesser extent, all companies are considering AI investments in their budgets, driven by their Boards of Directors, eager to take advantage of these new technologies.
There are at least three proven areas of impact that AI has within companies:
Productivity
This is where the vast majority of companies are currently focused. A recent Gartner study, conducted after interviewing more than 80,000 executives around the world, shows compelling results: it has enabled improvements of the order of 40% in collection processes, 98.5% in forecast accuracy, 90% in productivity in transactional processes, and millions of dollars in fraud prevention and risk monitoring.Decision-Making
From models in which AI makes decisions autonomously for humans, to a more traditional use of information as a basis for decision-making, to “augmentation” models where AI proposes different scenarios and humans decide.
So, the question we must ask ourselves is: How much power do we give AI for decision-making?Disruptive Capabilities
AI can penetrate the core of the organization and propose new business lines and models. To achieve this, the joint use of two main tools will be vital, both of which are worth briefly describing:Machine Learning, which, through processing massive volumes of data, can identify “normal” (patterns) and “abnormal” (exceptions to the pattern) behaviors, providing the ability to suggest the best “next step” with a high degree of certainty.
Generative AI, whose ability to generate content such as text, videos, and images, and which, powered by Machine Learning, will allow AI to fully enter the most core aspects of the business. Some proven examples worth mentioning are dynamic pricing applied in many businesses and robotic education.
We are moving into a quite uncertain world. Traditional technologies have so far supported humans, but generally have not had the goal of replacing them. AI proposes a concept of human replacement, whether partial or total, which will inevitably lead us to rethink the role of people in organizations.
What Should the CFO Pay Attention To?
I identify three main aspects to be taken into account:
1) Required Investments
The main problem we face is that the true cost of AI is still largely unknown. I propose to analyze the AI cost structure in three main sections:
Project Rollout: Consisting of preparation, implementation, and post-go-live support of the different projects.
Experimentation: This phase, based on trial and error, seeks to generate value opportunities for the organization through AI.
The Ongoing Phase: This includes the daily management of the implemented models.
Experience to date indicates that the rollout costs of AI projects, given the practice acquired by companies during the successive technological changes of recent decades, will prove to be a reasonably predictable and manageable factor.
This is not the case with the experimentation costs required by AI, which, due to the low level of maturity reached to date, must be estimated as accurately as possible and with frequent and precise monitoring. It is assumed that over time, experimentation and continuous learning will lead to this cost line adopting a downward trend.
Regarding the ongoing aspect, it is already clear that costs will be increasing; data models, management of complex languages and technologies, compliance assurance, etc., will require constant reviews. On the other hand, it remains to be seen what the pricing models of AI providers will be. Will we pay per employee/query/use/employee…? How many users will be given access? Undoubtedly, costs in this area will tend to increase, also considering that AI models will cover more and more functions of the organization.
The CFO must unmask the costs of AI as a way to prevent cost overruns. To do so, they must be involved in all phases, review agreements with suppliers, and monitor internal use to avoid avoidable contingencies.
Similarly, it is highly recommended that organizations manage projects with project portfolios, and it is there that the CFO and their teams must ensure that projects have a centralized governance framework aligned with the company’s strategies. An organization where each sector is experimenting with its own AI models without central coordination and the certainty that investments are perfectly aligned with business objectives would be undesirable.
2) Taking Care of the Company Reputation
Another key aspect to be taken into account is how to protect the company reputation.
Since the communication of business strategies and the actions developed for their execution is so important to the business community—after perhaps decades of work to achieve a good reputation—let’s not allow AI to ruin it for us.
AI has the ability to become the face of our organization with banks, clients, suppliers, investors, etc. Unreliable data, malfunctioning models, and human errors could eventually cause serious reputational damage.
The CFO’s role will be to identify risks and generate mitigation strategies before AI models come into contact with the organization’s stakeholders.
3) People Care
There is a clear fear among people, as it is widely accepted that AI will very likely replace a large part of human jobs. While this is partially true, it is also true that the role of people will need to transform—and that AI will require people to make the difference.
We know that certain tasks will disappear from our desks, but human intervention will have to continue in other dimensions of value that we will discover along the way.
It’s essential to make people understand this: a negative predisposition toward these changes will undoubtedly create a barrier not only to the speed but also to the quality of AI application in organizations. If this is the case, projects will fail.
The challenge in getting people on board with this wave of change will be knowing how to identify and communicate what is no longer being done and what you’re going to do next.
This will be a massive problem, affecting the entire company and all organizations.
This is why the CFO and their frontline staff are expected to be up to the task—learning and communicating, getting involved from the beginning, and setting the pace, faster or slower depending on each case.
We must always remember that if we all follow what AI tells us, literally, we will all be using the same data source and adopting the same ideas, so there will be no differentiation.
On the other hand, AI reserves for humans the possibility of working above AI by incorporating innovation concepts that can only be developed from a human mind.
I hope these thoughts help our colleagues understand how to engage and lead the incorporation of AI into the organizations.