AI in Higher Education How to Adopt and Adapt

Artificial intelligence is disrupting higher education, offering new pedagogical perspectives but also raising challenges. During a webinar organized by the G9+ Institute and EdTech France, experts and industry stakeholders explored the challenges of adopting AI in institutions, while sharing solutions to better support students and teachers.

Higher education institutions have no choice but to consider the rise of AI. Incoming cohorts are already familiar with online courses and artificial intelligence. Failing to prioritize this new technology limits student development and risks issues with engagement and retention,” says Renaud Monnet, director of the Digital Lab at CentraleSupélec.

Missing out on the AI revolution means losing touch with contemporary issues. This technology can also simplify tasks for teachers and enhance student experiences.

Mastering it, however, “takes time and requires a thorough approach ,” says Claire Lecocq , deputy director of the École Polytechnique information intelligence engineering school . This is where edtechs come in, offering digital tools to support these stakeholders in their daily practice.

“Higher education represents a spectrum of up to eight years of study, cohorts that are not the same size (they are large in bachelor’s degrees, smaller in master’s degrees, and monitoring is individual during the doctorate), with different facilitation challenges. The main problem for the teacher is being able to focus on the students who need the most support. AI can help to coach quickly and better ,” believes Renaud Monnet.

Claire Lecocq adds, “Many AI tools foster soft skills—a teaching aspect not achievable in lecture halls. These tools create safe, individualized spaces that elevate student experience.”

At CentraleSupélec, training programs for job interviews and public speaking are being tested. They provide emotional challenges for students and provide advice on how to improve based on a video, before moving on to a human and educational debriefing.

Generative AI’s proposals are not without their flaws and hallucinations. “We will need to consolidate our skill sets for critical thinking ,” notes Claire Lecocq.

But this weakness can also be a source of learning: “In mathematics, I use ChatGPT to create reasoning and I ask students to detect the true from the false. The teacher can use biases to warn and then use them as a basis to propose something better ,” illustrates Lucie Jacquet-Malot , pilot of the Digital Demonstrator in Higher Education (Demoes) project at the University of Picardy Jules-Verne .

Some edtechs, like Nolej , help get around this problem. “We chose not to start from a blank page: the teacher must select a resource. We have to keep the human in the loop ,” shares Nejma Belkhdim , co-founder and sales director of the edtech that allows the generation of interactive educational activities (quizzes, videos, fill-in-the-blank text, etc.) from statistical documents using AI.

Generative AI, however, may not last forever. “Today, there is uncertainty about what will become of generative AI in terms of reliability, but there are scenarios where it will improve. I am optimistic: let’s project ourselves into the hypothesis where we will no longer have to be wary ,” suggests Renaud Monnet.

How can we convince teachers to embrace AI? “By leveraging the power of certain edtechs to show them everything we can do,” suggests Lucie Jacquet-Malot. “The teacher has to see a benefit in it: saving time or helping students develop skills they can’t always assess.”

This is what the University of Picardie Lorraine experienced with Nolej. “We observed total support from teachers without any need for compensation ,” continues Lucie Jacquet-Malot.

Cooperation between institutions and edtechs allows for faster progress and better teacher buy-in, adds Nejma Belkhdim.

Edtechs can support the culture of innovation in institutions. And universities can support us too. We are currently reworking the entire technical architecture of Nolej to make it easier for universities to plug in, which is only possible if institutions play the game by investing time and money ,” she explains.

The educational and technical issues are the same internationally, notes Nejma Belkhdim. She therefore encourages edtechs to “look beyond France and demonstrate speed . “

The co-founder of Nolej, however, raises difficulties specific to higher education: the slowness of procedures and the threat to the sustainability of projects funded for a fixed period (as is the case with calls for projects launched at the national level.

Within the Digital Lab, CentraleSupélec is developing its own AI tool, Aristote . This initiative allows the community to be involved, particularly because the prototyping system is student-centered. “Working together to improve the teaching approach creates a dynamic between teachers and students ,” says Rennaud Monnet.

This tool aims to address the intellectual property issues raised by ChatGPT. Indeed, the lack of transparency regarding the use of American AI data makes teachers, researchers, and even businesses wary of sharing their content.

Aristotle also illustrates the cooperation between edtech and higher education, as it can be connected to edtech solutions, ultimately allowing the user to choose their AI solution.

With the advent of ChatGPT, “the governance of higher education institutions and teachers have faced a very rapid paradigm shift ,” recalls Lucie Jacquet-Malot. Initially, it was not necessary to support teachers who took hold of the tool, but what takes longer to put in place is a real institutional strategy. Projects like New University Curricula (NCU), Demoes and Skills and Jobs of the Future (CMA) are an opportunity to put in place governance strategies that offer a different perspective on AI than that of an average user.”

She continues Without strong support from the vice-presidents for digital and the training and university life committee, it will be complicated. They are in the committees, working on creating models, etc.

To speed up the university’s purchasing phase, which, as a public institution, must issue a call for tenders before choosing the solution to adopt, it is possible to use innovative purchasing. This system allows buyers to award innovative works, supplies, or services contracts worth less than €100,000 excluding tax, without prior advertising or competition.