PTThe professor and their digital intern
How, and for what, do business administration professors use generative artificial intelligence? Thirteen professors described what they gain, what they distrust and how their own role is changing. Explained in 14 screens.
Would you let an intern prepare your class?
One of the professors interviewed put it this way: “AI is the professor’s assistant. And like all of them, an intern, let’s say, right? […] I have to train it.”
Imagine you are a professor. For each task, choose what you would do with this intern. Then see what the interviewees said.
Illustrative exercise created for this page. The quotes come from professors interviewed for the study (Section 5), identified only by interview number, and are translated from Portuguese.
AI has already entered the staff room
Since ChatGPT was launched in November 2022, generative AI has spread through schools and universities. Recent surveys cited in the dissertation show how far it has gone.
Cards: survey by Fundação Itaú (a Brazilian education foundation, 2025) of students, teachers and school administrators, cited in Section 1.3.1. Not data from this dissertation.
Chart: survey by UFMG (Federal University of Minas Gerais; Silva et al., 2025) of the 410 faculty members who use AI, cited in Chapter 5. Respondents could tick more than one option. Not data from this dissertation.
Hardly anyone had asked business administration professors
The author searched two large databases of scientific articles (Web of Science and Scopus) for studies on AI and professors in business administration programs. Hover over each step.
- 102
results across the two databases (51 in each)
- 34
articles from November 2022 onward, when ChatGPT was launched (after filtering by type and language)
- 23
distinct articles, not counting those that appeared in both databases
- 15
had all the search terms in the title, abstract or keywords
- 3
actually dealt with how business administration professors use AI
Most studies look at students: how they use AI in exams and assignments and what it does to their learning. The professor’s side was left almost unanswered.
Source: Section 3.4. Disc areas are proportional to the number of articles.
13 professors, 8 institutions, one long conversation
The author interviewed professors from undergraduate business administration programs, with open questions about how they use AI, for what, what they fear and what has changed in the way they teach. She stopped when the answers began to repeat themselves.
There were 14 interviews. The first was left out of the analysis because that professor no longer taught in the business administration program. Source: Sections 3.1 and 3.3, Tables 1 and 5. The median age (58) is our calculation.
From 65 keywords to 4 ideas
The interviews were transcribed and read in search of patterns. Five keywords came out of each one, backed by excerpts from what was said. They were grouped into codes, the codes into themes, and the themes became four concepts. Click a layer.
Method: systematic thematic analysis (Naeem et al., 2023), carried out with the help of the AI tool NotebookLM and reviewed by the author. Source: Chapters 4 and 5. The total of 65 keywords (5 × 13) is our count.
“Saving time” comes up most often
Each row is a professor; each column, a topic. 11 of the 13 spoke of saving time, speed or productivity. Close behind come AI’s mistakes and the fear of depending on it. Click a square to read the quote.
Our count and grouping, based on the keywords the author selected for each interview (Section 5). Keywords and quotes are translated from Portuguese; the counts refer to the original interviews. The author states that time stood out in every interview.
A case study that took days now takes 15 minutes
One professor said she used to write a case study “in 2, 3 days. Now I do it in 15 minutes”. The study calls this catalysis of efficiency: operational tasks move faster, leaving more time for work that demands attention.
Scale in calendar days: 15 minutes take up about 1% of a day. The quote does not say how many working hours went into each day. For professors who teach at more than one institution, the time saved matters even more (Section 7). Source: Interview 14 and Theme 1.
The time saved comes from very concrete uses, almost all of them tied to teaching rather than research. Click each one.
…but nobody trusts it 100%
The same professors who save time distrust what AI delivers and fear what it does to thinking, both theirs and their students’. The study calls this technological cognitive vulnerability. The professors explained it with everyday images: click each card.
By our count, 8 of the 13 spoke of mistakes or of the need to check everything, and 8 of the risk of depending on the tool. To protect themselves, some only use AI on subjects they know very well (Interview 14). Source: Section 5 and Theme 2.
From keeper of the truth to “safe link”
With AI in the classroom, the relationship is no longer just between professor and student: a third element appears. For the interviewees, the professor stops being the one who hands down content and becomes the one who guides, checks and makes sense of what the machine produces.
The study calls this shift the mediating redefinition of the teaching role. The opposite fear also came up: one professor thinks AI may reduce the number of teaching jobs (Interview 4). Source: Theme 3 and Section 7.
Four ideas that fit together
At the base, what AI is used for: assistant and accelerator. In the middle, the tension this use creates. At the top, the response that, according to the author, could overcome it: the professor as mediator. Click each concept.
Our drawing, based on the text of the conceptualization and the conclusion: the dissertation describes the four concepts and how they relate, but does not include a figure of the model. Theoretical basis: Donald Schön’s epistemology of practice (1983).
If AI disappeared tomorrow, almost nothing would change
The professors acknowledge that AI is a “road with no way back”. Even so, when the author asked how they would feel if it did not exist, only one professor said they would feel different. The others would feel as they always had.
The dissertation does not identify who that professor was. Source: Chapter 5.
Reflection exists, but it is still shallow
AI led the professors to rethink their practice, but the author interprets this reflection as happening “in a superficial and pragmatic way”: the focus is on keeping up with demands and “not falling behind”. Drag the marker to explore each end of the scale.
The scale is our metaphor. Its two ends, superficial and emancipatory, come from authors discussed in the dissertation (Brito, 2006; Reali and Reyes, 2009). Source: Sections 7 and 8.
So what now?
Business administration is one of the degree programs with the most students and graduates in Brazil. The way its professors deal with AI reaches a lot of people, and employers expect both professors and students to know how to use it.
For universities and colleges
Many interviewees said their institutions still have no rules on the use of AI. Without guidelines, each person decides alone, and Brazil’s own Ministry of Education warns that this can deepen inequalities and create ethical problems.
Dissertation, Chapter 5For professors
The interviewees themselves offer the practical rules: treat AI as an intern that needs supervision, review everything and use it only on content you know well. In class, it can become an “instrument of critical alert” to show students where the machine gets things wrong.
Interviewees’ words · our selectionFor students
AI is “a means and not an end”. The fear professors repeated most is that students will swap reasoning for the tool. Using AI without losing the ability to think critically is the skill they hope to develop.
Our reading of Theme 2For business administration programs
Programs tend to prioritize “learning by doing”. For reflection on AI to go beyond the pragmatic, the author suggests that the education system give professors time and space to think about their own practice, not just courses on tools.
Dissertation discussion (Section 7)What this study does not say
- It is a portrait of 13 people, not of the country. Qualitative research seeks depth, not representativeness: 12 of the 13 professors work in São Paulo state.
- A specific generation. The interviewees are 45 to 71 years old and trained before generative AI existed. Younger professors may see the tool differently.
- A concentrated profile. 10 of the 13 are men and 11 of the 13 teach at private institutions (our observation from Table 5).
- It is what professors say. The study gathers perceptions reported in interviews, not observation of classes.
- Suggested next steps: hear from professors in other regions and in countries without AI guidelines, and understand how AI could prompt deep reflection on practice, not just pragmatic reflection.
SÁ, Adriana Fontes. Inteligência artificial no ensino superior: uso pelos docentes do curso de Administração [Artificial intelligence in higher education: use by faculty in the Business Administration program]. Master’s dissertation (Master’s in Business Administration) — Graduate Program in Business Administration, Mackenzie Presbyterian University, São Paulo, 2026. Advisor: Prof. Dr. Diógenes de Souza Bido.
Qualitative, exploratory research based on 13 semi-structured interviews examined through thematic analysis. Examining committee: Prof. Dr. Diógenes de Souza Bido, Prof. Dr. Janette Brunstein (Mackenzie) and Prof. Dr. Dirceu da Silva (Unicamp, the State University of Campinas).
This interactive version simplifies the work for a general audience; the theoretical framework, the interview guide and the full quotes are in the original text (in Portuguese).