From Adoption to Embrace of Generative AI in the Workplace: A Dialogue Between Educational Sciences and Psychology

Chapter 1: AI Uses and Adoption: A Deconstructed View of Work

July 10, 2026

The arrival of generative artificial intelligence in the workplace raises many questions. While discussions often focus on productivity gains or changes in job roles, they sometimes overlook a crucial question: What actually happens when these tools are integrated into work practices?

The rise of generative artificial intelligence marks a new stage in the transformation of the world of work. Beyond productivity gains and the automation of certain tasks, its integration raises deeper questions about the evolution of professional practices, skills, and ways of collaborating. While the adoption of these tools is widely emphasized today, it is not enough to grasp the true changes taking place within organizations. We must therefore ask ourselves what has really changed in our relationship to work and consider the organizational challenges that go beyond the mere issue of technology adoption.

This first interview is part of that effort and is structured around several questions designed to explore these issues. By combining perspectives from the fields of education and psychology, they examine the conditions under which generative AI is integrated into the workplace, the learning it requires, and the questions it raises regarding skills, professional identity, motivation, and occupational health.

Since generative AI has been adopted in professional settings, what do you think has changed the most in terms of our relationship with work?

Joseph Baud-Grasset: What strikes me most about the emergence of generative artificial intelligence is that it has become established in professional practices even before organizations have really had time to consider its implications. I’ve observed that professionals have begun using tools like ChatGPT to create content, draft documents, or structure projects long before any usage guidelines or institutional recommendations were in place. Faced with these new “cognitive companions,” professionals are increasingly becoming reflective practitioners, as defined by Donald Schön (1983). Professionals engage inreflection-in-action andreflection-on-action.

Thus, in my view, this shift reveals a transformation that originates first and foremost in practices rather than in institutions. Its spread is happening at an unprecedented pace: whereas certain changes used to take several years, just a few months are now enough to alter professional practices and attitudes.

In the field of training, the time savings are undeniable. Designing an initial module, creating instructional activities, or producing training materials can now be done more quickly. However, the issue is not limited to automation. What is undergoing a profound change is the nature of the work itself. The time previously spent researching, writing, and organizing also served as a period for reflection and assimilation. Now, I observe that professionals are increasingly acting as evaluators, arbiters, and critical designers of content they did not entirely produce themselves. While certain tasks have become less burdensome, the demands for discernment, validation, and accountability have intensified.

Skills and approaches are becoming increasingly complex as generative AI pushes professionals toward new areas of expertise. Indeed, this shift sometimes fuels the idea that training projects should be completed ever more quickly. However, while artificial intelligence speeds up certain stages, it does not replace needs analysis, the quality of instructional design, or the ability to adapt training to specific contexts and individuals.

This leads me to believe that knowing how to ask questions of artificial intelligence, assess the relevance of its responses, identify its limitations, and decide when not to use it is now an essential professional skill. More than just technical proficiency, it is a matter of learning to exercise discernment.

Stéphane Bonzon: Joseph describes what is accelerating. However, in addition to the technological acceleration brought about by generative artificial intelligence, I would emphasize the acceleration of social change, as defined by Rosa (2010)—that is, the rapid shift in expectations and reference points. In the workplace, this leads to a shift in norms without them having been renegotiated or explained. For professionals, this creates a strange experience: we continue to do our work, but we sense around us that the implicit standard is shifting. The long-term development of expertise built up year after year through diverse experiences and encounters—the time it takes to develop a professional signature, a “style,” or a reputation patiently forged—no longer moves at the same pace as technical time. Yet no one says this explicitly. We therefore find ourselves in a gray area, where just about any content—for example, a summary note produced in five minutes using AI—can replace a piece of work that used to take three hours. And the problem is that we’re no longer quite sure whether our version is still better or merely equivalent, nor whether the difference in quality we champion is still recognized—or even simply perceived—by anyone.

There’s a lot of talk about AI adoption. Is that enough to understand what’s really going on within organizations?

Joseph Baud-Grasset: When organizations talk about adopting AI, they often focus first and foremost on deployment: purchasing licenses, granting access, offering training, or organizing a demonstration.

Yet in adult education, we have long known that just because a tool is available doesn’t mean it’s appropriate. Learning doesn’t begin simply because a resource exists; it begins when a person finds meaning in using it within their own activities.

Simply deploying the tool is not enough. We must also create the conditions that will allow professionals to integrate it into their work.

In many cases, it is shared spaces that make adoption possible. Thus, taking adoption seriously means accepting that the transformation of practices will be slower than the rollout of tools. The most advanced organizations will not necessarily be those that have adopted AI the fastest, but those that have created the conditions for its integration: time, trust, spaces for exchange, and the opportunity to experiment without excessive pressure. Making room for the creation and establishment of communities of practice, as defined by Etienne Wenger (1998), seems to me essential to the integration of demanding, evolving expertise. For the challenge is not merely to use generative AI, but to develop a use case that makes sense in the context of real-world work and its perpetual evolution.

Stéphane Bonzon: The debate on the adoption of AI in business is implicitly structured by the UTAUT model (Venkatesh et al., 2003), as revised by Blut et al. (2022). The four main determinants of intention and behavior regarding the use of a technology in the original model are performance expectancy, effort expectancy, social influence, and facilitating conditions. These variables are useful for explaining why a person uses a tool, why they state their intention to use it, and why their organization promotes or hinders adoption. This can predict usage but not what that usage does to the individual. This is precisely the distinction Joseph makes when he differentiates between deployment and appropriation. One can check all the boxes in the UTAUT model and still have achieved only behavioral adoption. Appropriation, in the strongest sense, is something else entirely. The study by Baruel Bencherqui et al. (2025) on thirty-two skilled employees, which draws on the UTAUT theoretical model, categorizes professionals’ attitudes toward generative AI into three distinct profiles, reflecting a disparity among users. A few pioneers use it extensively and communicate extensively about it, but the majority continue to use it only occasionally and cautiously. Some, held back by fears regarding data security or a lack of control, simply do not use it. The relationship professionals build with generative AI strongly influences their decision to use it or not. As Rabardel (1995) demonstrated with the concept of instrumental genesis, a professional does not simply receive a finished technical system that they use as is. On the contrary, they develop their own work tool by transforming both the tool itself and their own activity. From this perspective, adoption is the moment when AI enters the work environment, whereas appropriation begins when the user transforms it into a tool relevant to their work. However, with a probabilistic system such as generative AI, the relevance of its use is never self-evident. In my view, only an informed appropriation of generative AI—as I like to call it—can prevent it from becoming a mere cognitive crutch or an automatic delegation mechanism.

Resources:

Baruel Bencherqui, D., Chaari, J., Le Flanchec, A., & Mullenbach Servayre, A. (2025). The Adoption of Generative Artificial Intelligence (GAI) by Employees in the Workplace. Vie & Sciences de l’Entreprise, 224(1), 13–40. https://doi.org/10.3917/vse.224.0013
Blut, M., Chong, A. Y. L., Tsigna, Z., & Venkatesh, V. (2022). Meta-analysis of the Unified Theory of Acceptance and Use of Technology (UTAUT): Challenging its validity and charting a research agenda in the red ocean. Journal of the Association for Information Systems, 23(1), 13–95. https://doi.org/10.17705/1jais.00719
Bonzon, S. (April 10, 2026). Generative AI and Work: Beyond the Hype, What Research Already Reveals. ISFB Observatory on Banking and Financial Skills.
Rabardel, P. (1995). People and Technology: A Cognitive Approach to Contemporary Instruments. Armand Colin.
Rosa, H. (2010). Acceleration: A Social Critique of Time (trans. D. Renault). La Découverte. (Original work published in 2005)
Schön, D. A. (1983). The Reflective Practitioner: How Professionals Think in Action. Basic Books.
Wenger, E. (1998). Communities of Practice: Learning, Meaning, and Identity. Cambridge University Press.

Short AI Training Course for Finance Professionals

© Institut Supérieur de Formation Bancaire (ISFB). All rights reserved.
The analyses and content published by the ISFB may be quoted or reproduced in part, provided that the source is clearly mentioned. Any full or substantial reproduction of this article in another medium or format is subject to the prior written authorization of the ISFB. In order to facilitate reading and without any intention of discrimination, the masculine gender is generally used, in accordance with the grammatical rule that allows it to be used as a neutral value 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.

Joseph Baud-Grasset

Training Coordinator
ISFB

Stéphane Bonzon

Career Counselor
ISFB

Knowing how to ask questions of artificial intelligence, evaluate its responses, and decide when not to use it is becoming a professional skill.

Joseph Baud-Grasset