02.07.2026, Moscow.
Artificial intelligence has become a major challenge for the entire educational system: previous forms of knowledge assessment are no longer effective, while new ones have not yet been developed, Rossa Primavera News Agency‘s Social Desk notes.
Traditionally, written assignments – ranging from essays and term papers to theses – were considered effective ways to test students’ knowledge. However, this is no longer the case, because written assignments can be created using neural networks in literally seconds, and students are actively using them.
Moreover, such AI plagiarism is difficult to detect. Existing AI-text detectors are imperfect, they often malfunction and are not suitable for making important academic decisions. Furthermore, these tools are not difficult to “trick” using simple methods, such as paraphrasing. In addition, due to the unreliability of detectors, there is a danger of false accusations, when a written assignment independently completed by a student is mistakenly identified by the system as having been created with the help of a neural network.
On June 29 at the forum AI Technologies in Higher Education, the Russian Minister of Science and Higher Education Valery Falkov drew attention to the problem of assessing students’ knowledge, stating that “theses no longer fulfill the function they performed in higher education 10-15-20 years ago.” He noted the need to replace the previous text-based format of student work.
But what is proposed as an alternative? Experts are calling for shifting the focus of assessment from the final written product to the learning process itself. Oral exams, live presentations, project work, and their defense are being considered as alternatives to written assignments. For example, in addition to submitting a written term paper, an oral defense could be introduced, during which the student would be required to explain their reasoning and justify their conclusions. Their ability to engage in reasoned argument, as well as to verify and analyze facts, could also be assessed.
In principle, there is nothing new in these formats, but it is proposed to shift the emphasis precisely onto them. Doing so is not so easy. The higher educational system is quite inert; it cannot quickly restructure the entire knowledge assessment system developed over decades, nor change the very approach to it.
The mass use of AI by students is forcing the entire system to rethink itself. Obviously, neural networks should not replace the ability to think independently. But assessing this ability in students has now become a major problem, for which a systemic solution has not yet been found.
Source: Rossa Primavera News Agency
