The debate over artificial intelligence in higher education has become impossible to ignore. What began as heated exchanges in journalism seminars has evolved into a broader reckoning across French universities and business schools, where faculty and students alike are grappling with a technology that promises to transform their professions—or render them obsolete.
An investigation into how several major institutions are responding reveals four key lessons about the state of AI adaptation in French higher education.
Students Are Neither For Nor Against—They’re Everywhere at Once
There is no single student stance on AI. Usage has become pervasive and routine. At Sciences Po’s journalism school, AI has even replaced Google as the default reflex for students who, according to director Alice Antheaume, “operate in prompt mode.”
At ESSEC business school, more than half of students use AI daily, with another third using it weekly. Estefania Santacreu-Vasut, associate dean for pedagogy, describes their relationship with the technology as “quite realistic”—neither overly enthusiastic nor dismissive.
What varies is where anxiety takes root. At Rubika, a Valenciennes school specializing in animation, video games, and design, resistance maps precisely onto the professions most under threat: massive rejection in animation, pragmatic acceptance in industrial design, and a finer split in video games, where junior programmers feel the pressure while artists remain largely unthreatened.
At Grenoble’s Ensimag engineering school, which trains the very designers of AI tools, Christophe Picard observes “the entire spectrum,” from hyper-enthusiasts to students who are “downright opposed.” Notably, the most heated debates occur not in computer science courses but in interdisciplinary classes—where the focus shifts from what AI does to jobs to what it does to society.
At Sciences Po Journalism, Antheaume notes that “the resistance that exists is not ideological but ecological, driven by the carbon footprint of data centers rather than distrust of the tool itself.”
When rejection takes concrete form, it can be dramatic. When Rubika animation students learned that the Chamber of Commerce and Industry was hosting an AI event in their building, they showed up en masse to protest what they saw as celebrating a direct threat to their future careers.
Teaching Differently: Faculty Under Pressure
For professors, the transformation is quieter but equally profound. Each institution has developed its own approach, and these approaches differ more than one might expect.
Rubika has chosen a temporal strategy: the first three years of study deliberately limit AI to ensure mastery of fundamental skills, with integration coming later, validated by annual surveys of alumni to confirm real-world utility in studios.
Forward College has opted for spatial separation: a locked, AI-free zone for assessments, and a space where AI is not only tolerated but encouraged—provided students maintain a log of their strategy.
ESSEC delegates the decision to individual professors. Its institutional policy, adopted this summer, allows each teacher to define, course by course, their “technology-free zones,” activities where AI is encouraged, and those where it is merely tolerated.
The challenge is preserving the teaching foundation. At HEC, Anne Michaut identifies the top priority as bringing along a faculty with vastly unequal levels of technological appetite—without ever letting a professor believe their course has become obsolete.
Students themselves want a voice. Santacreu-Vasut notes that three-quarters of ESSEC students want to be involved in institutional decisions regarding AI.
Grading: The Marker of Added Value
The first casualty of this technological upheaval is already clear: homework. And so is the consequence—a renewed investment in what happens in the classroom.
Sciences Po Journalism has shifted 70% of its grading toward in-person work. At Ensimag, many instructors have abandoned take-home assignments in favor of final exams on paper or in strictly controlled environments. Some prefer a hybrid approach: free AI use during the semester, followed by an isolated final session on a slightly modified topic to verify students can reproduce their learning independently.
Forward College simply locks internet connections during assessments. ESSEC goes further, requiring oral examinations with no technological assistance to ensure students can defend and argue their work alone.
The goal is not to punish cheaters but to ensure AI remains merely a tool.
Training for Jobs That May Not Exist
The most significant challenge is preparing students for a job market that has yet to determine how AI will reshape it. Students understand that the employment landscape has frozen, though whether this is entirely due to AI remains unclear.
Behind declining hiring and multiplying layoffs, two factors emerge:
- Poor global economic health, with AI serving as a convenient pretext for companies seeking cost savings
- Boards instructing hiring freezes based on assumptions that AI will replace many positions—a scenario that remains far from reality
Alumni invited to share their career paths sometimes put current difficulties in perspective, reminding students that this is not the first technological revolution they have faced.
Whether AI is to blame or not, the impact on these schools is concrete: declining job placement rates. Ensimag, which historically reported over 95% placement within a year, has dropped to 85% in a single year. At Rubika, where external audits have tracked placement for three years, the historical rate of around 90% fell to 85% last year in animation, with further deterioration observed this year. In video games, the decline is even more pronounced in an industry Rubika describes as “devastated” for over a year.
The Bonus Lesson: Fractures of (Non)Adaptation
Perhaps the most counterintuitive finding is that schools are not describing a clean technological rupture with clear winners and losers. Instead, they describe permanent adjustment, sector by sector, where AI often serves less as a cause than as a mirror—magnifying existing fractures between professions, between generations of teachers, and between technophile and technophobe students.
Another fracture is emerging between those who move forward and those who wait. Every school contacted for this investigation had engaged with the issue—except one. The Regional Professional Training Center for Lawyers in Marseille has not yet addressed AI in its curriculum. It may do so next year.
One strong conclusion emerges: mastery or nothing. Those who master AI, maintain critical thinking, and develop their creativity will fare better than others. The question becomes: where does each person’s added value lie?
It sounds simple. It is not. Everything pushes toward absorbing this technological revolution without asking questions.
Without a comprehensive overhaul, fractures will widen. As Boris Walbaum of Forward College emphasizes, what is at stake is the appetite to learn. “And the appetite to learn is human interaction.” This means looking further ahead—treating AI not as a separate subject but as part of a whole, rethinking the entire educational chain, and better identifying the skills future generations will need in the very near future.
