As AI transforms higher education, faculty experts discuss what comes next for research in the Arab world
When we think about AI in academia, the typical image is of a rushed student using a large language model (LLM) to generate an essay right before the clock strikes 11:59 pm. Since ChatGPT’s release in 2022, artificial intelligence has become both a challenge and a tool for educators worldwide. Yet beyond the classroom, the technology is also transforming academic research, creating new opportunities while raising questions about bias, authorship and academic integrity.
How is AI reshaping research across the Arab world? AUCToday spoke with two faculty approaching the question from different standpoints. Nagla Rizk ’83, ’87, professor of economics and founding director of AUC’s Access to Knowledge for Development Center, brings an economic and development perspective to questions of AI bias and inclusion through her work at the center, while Adam Talib, associate professor in the Sheikh Hassan Abbas Sharbatly Department of Arab and Islamic Civilizations and a scholar of premodern Arabic literature, examines AI through the lens of language, literature and culture.
How is AI changing academic research in the MENA region?
Talib: LLMs can democratize access to historical primary sources that have long been inaccessible because of language barriers. Imagine the average Egyptian student interested in Andalusi literature. Traditionally, the answer would have been: You need to learn French, and maybe Spanish, before you can fully engage with the scholarship. Now that may not be as pressing because an LLM can help you get the gist of a French article. It pushes language learning in a lateral direction. It doesn’t eliminate the need to learn the language, but I think it will encourage people to be a little bit more broad-minded in how they approach their training.
How can organizations protect regional research from marginalization in the AI boom?
Rizk: AI output depends on what is being fed into the system as data to train its models. If regional research output is not integrated into AI platforms, it will be invisible to AI users. For this reason, among others, technology has a tendency to be unjust, so you need to integrate justice into its design. You need to make sure diverse knowledge content and languages receive equal attention so no one is left behind. Regional research has to be included in datasets, and regional researchers have to be at the forefront of shaping the technology. But this is not easy. Who takes care of this? Who is responsible? We all are. Technology needs to be inclusive, and that is everyone’s job — governments, researchers, businesses, civil society and international organizations.
"Technology has a tendency to be unjust, so you need to integrate justice into its design"
How is AI changing student learning?
Talib: AI risks diminishing the rigorous training that teaches students how to think and engage with resources critically. AI companies talk about the amount of hours they put into training their models. The average MA or PhD student also spends thousands of hours developing that same depth in knowledge. The challenge will be motivating students to see the value of that training, despite the fact that somebody with no Arabic can sit in their house at 11 pm and start reading Arabic medieval chronicles with the help of AI. How do I persuade an intelligent student to put in the work to read it themselves? You have to show them that a specialist using an LLM is an order of magnitude better than an LLM on its own.
Rizk: Students will use AI even in contexts where access to technology is limited, and such use should be responsible. Ideally, students should be allowed to use AI in real time during class and compare what they get with insights shared in their earlier offline discussions. One way to avoid excessive student use of AI in teaching and learning would be to rely more on oral and in-class exams with no access to the internet or phones. We are fortunate at AUC, as we can afford all this. We have the facilities, and we teach small groups of students. This is not the case for other parts of Egypt. Think of universities in less privileged areas with fewer resources and much larger class sizes. How can they adopt such personalized approaches? That’s a real challenge, and it is one that AUC has helped address through its Statement on the Use of Artificial Intelligence Tools.