Back to pen and paper and smartphone ban: universities are forced to change the format of classes and tests, otherwise AI prevents both teaching and assessment
Artificial intelligence is forcing leading universities around the world to reconsider familiar forms of teaching and assessment. They are starting to abandon essays, "home projects", and other assignments that students complete outside the classroom. Teachers are returning to in-class, handwritten tests, replacing some tests with oral exams. At the same time, some universities are banning laptops and smartphones in auditoriums.

The Massachusetts Institute of Technology is considering changing its entire education system in response to the emergence of powerful artificial intelligence models. This is discussed in a new report published by the university's special committee on artificial intelligence. The publication Inside Higher Ed draws attention to it.
The report's authors came to an unpleasant conclusion for the education system: modern AI is already capable of providing convincing solutions for almost any written assignment in a bachelor's program — from essays and program code to mathematical problems and proofs.
For instructors, this problem is not new: with the spread of generative AI, it is increasingly difficult to determine whether a submitted essay, a solved problem, or written code demonstrates the student's own knowledge.
The university cannot control how students complete work at home. Therefore, methods of knowledge assessment are also changing. Instructors are increasingly turning to methods that might have seemed outdated until recently: oral exams, handwritten assignments in the classroom, in-class discussions, and practical tasks.
This approach is also practiced in other American universities. As the Associated Press writes, oral defense of written assignments is gaining increasing popularity.
For example, at Cornell University, students on some courses, after submitting homework, undergo a 20-minute conversation with the instructor or their assistants. In this case, it is no longer the written work itself that is assessed, but rather how well the student understands what they have done and can explain the solution process.
Oral assessment is also being adapted for large groups. At the same Cornell University, a traditional final exam in one course was replaced with a half-hour conversation with each student, and an engineering course with 180 students conducts short, approximately four-minute individual interviews. The university is already organizing separate seminars for instructors dedicated to oral forms of knowledge assessment.
At the University of Pennsylvania, some courses have also started combining written assignments with oral exams. A history lecturer at the University of Virginia told Nature that she offers students to write essays directly in the auditorium three times a semester. They can use books and their notes, but they write the text by hand in an examination booklet — without phones or laptops.
As a result, some of the notable educational "innovations" of the artificial intelligence era are methods known long before the advent of computers: handwritten assignments and oral conversations with an instructor.
For American higher education, the return to oral exams is a rather significant change. Unlike some European universities, they were not a common form of assessing undergraduate students' knowledge in the US. Interest in them began to grow during the pandemic, when the transition to distance learning exacerbated the problem of cheating, and after the appearance of ChatGPT in 2022, this process accelerated significantly.
No AI in the auditorium
The University of Chicago Law School went even further. In its strategy "Reimagining Legal Education in the Age of AI" published this summer, it announced an experiment that will begin with the new academic year.
One of its principles is "removing artificial intelligence from critical moments of the learning process," such as during exams or studying fundamental concepts in the classroom. As part of a pilot project, electronic devices, with some exceptions, will be banned from all core first-year courses.
Changes will also affect the preparation of the Significant Research Paper (SRP), which students complete as part of their studies. Now, each student will have to orally discuss their work — give a presentation to a group or meet personally with an instructor.
At the same time, the goal of the new strategy is not simply to limit the use of AI. The Law School wants students to master fundamental legal concepts, learn to think independently, rigorously, and creatively, and at the same time, master AI as a tool for their future professional work.
Similar changes are happening at Princeton University. In May 2026, faculty almost unanimously decided that in-class tests and exams should be proctored. Thus, the university abandoned a 133-year-old tradition associated with the Honor Code, adopted back in 1893. According to it, exams were conducted without instructors and observers: students pledged not to cheat and to report observed violations. The Honor Code itself continues to operate, but now observers are present in auditoriums during exams.
And in August, the dean responsible for undergraduate education at Princeton urged faculty to be more critical of the use of electronics in class. He reminded them that they can prohibit students from taking notes on laptops, demand that phones be put away or turned off, and even completely switch classes to an analog format by asking students to bring printed texts.
Why students need to go to university
But the consequences of AI's spread are not limited to knowledge assessment and class organization. The aforementioned MIT report notes that it is already changing the very social life of the university. Students attend fewer instructor consultations, participate less in online discussions, and, according to university staff observations, gather less frequently in study groups in dorms and libraries.
According to the authors of the MIT report, the problem is also that effective learning requires not only intellectual effort but also interaction with other people. A student learns when they, together with classmates, unravel a complex mathematical proof, repeatedly redo a failed experiment, or argue about a solution — instead of immediately getting a ready answer from AI.
"Learning works when it is both challenging and social; knowledge is built through cognitive friction," the report states. That is why, its authors emphasize, it is important for students to physically attend university.
The committee proposes creating more opportunities for in-person meetings on campus, introducing technology-free periods dedicated to live communication, and placing a greater emphasis on collaborative learning in the classroom. According to the authors, adapting education to the new reality will require the efforts of the entire university community — students, faculty, staff, and administration.
Comments
Пабольшэ спорта нада, сьвежага воздуха.