Levels of Cognitive Activity and AI: Where Assistance Ends and Substitution Begins

2026/11/06, 10:46
Generative artificial intelligence has ceased to be an exotic tool and has become an everyday “co-author” of student work.

Yet behind its apparent efficiency—instant answers, well-structured essays, and ready-made calculations—lies a subtle pedagogical trap. The main question today is no longer whether to ban neural networks or legitimize them, but how students’ thinking changes when routine and even analytical cognitive work is delegated to an algorithm. To answer this, we need to move beyond general discussions and examine human–AI interaction through the lens of cognitive levels.

The impact of generative AI on learning is uneven. At the levels of remembering and understanding, models perform effectively: they find definitions, adapt complex texts, and generate examples and analogies. Students who once spent hours searching for basic explanations can now get them in seconds. But when a condensed explanation from AI replaces engagement with a primary source, learning becomes superficial and the apparent grasp of the material illusory.

A qualitative shift occurs at the level of application, when students are required not merely to reproduce theory but to solve a practical problem. Consider a specific case: analyzing the impact of a change in the Federal Reserve’s key interest rate on the currency markets of EAEU countries. A text generated by a neural network can have a flawless structure, polished wording, and even correct conclusions. But if assessment is limited to reading the finished document, the instructor cannot tell whether the student has mastered the calculation method or simply copied someone else’s algorithm. In that case, no competence is developed. The situation changes when the assignment requires students to show their calculations in detail, justify their choice of methodology, and clearly indicate where the algorithm’s involvement begins and ends. AI then becomes not a substitute performer but a training tool for developing professional skills.

At the levels of analysis, evaluation, and creativity, the risks are greatest. AI speeds up data processing, offers counterarguments, and helps structure reasoning. But these are precisely the areas where students must demonstrate critical judgment, ethical responsibility, and original thinking. When AI is used unchecked at these levels, thinking becomes formulaic, reasoning superficial, and authorship blurred.

Developing professional competencies requires three cross-cutting skills. The first is AI literacy: the ability to distinguish a model’s output from fact, understand the tool’s limitations, and recognize the biases built into it. The second is self-regulated learning: the ability to plan which steps to use AI for and which to complete independently. The third is digital ethics and attribution: the willingness to disclose the extent of the algorithm’s involvement and take sole responsibility for the intellectual work produced.

Thus, the task of today’s economics instructor is not to build an impenetrable defense against neural networks, but to deliberately develop three key metacompetencies in students: AI literacy, self-regulation, and digital responsibility. Only through a carefully designed assessment system that focuses not on the flawlessness of the final document but on the research process, the rationale behind methodological choices, and cross-checking facts can we develop independent professionals rather than passive consumers of algorithmic answers. Such professionals can critically assess information, manage generative AI tools, and take full personal responsibility for the final analytical product.

Author: Candidate of Economic Sciences, Associate Professor, Department of World Economy and International Finance, Financial University under the Government of the Russian Federation, Elizaveta Valentinovna Ogloblina.

This material has been translated using AI-powered neural networks. If you spot any errors, please highlight them and press Ctrl+Enter or notify us at info@nationalcapital.in