TRANSFORMATION OF LEARNING TASKS USING GENERATIVE AI

Abstract

The article addresses the transformation of learning assignments in the context of the widespread availability of generative artificial intelligence (AI) tools in higher education. The relevance of the topic is driven by the contradiction between students' mass use of generative AI to complete assignments and the lack of tools for their didactic renewal. A common response among higher education institutions is to protect academic integrity by detecting and prohibiting the use of generative AI. However, this approach does not resolve the underlying contradiction, as an assignment that AI can complete in full reveals a didactic misalignment between the assignment itself and the learning outcomes it is meant to achieve. The purpose of the article is to systematize the types of learning assignment transformation in the context of generative AI and to present their practical implementation in the form of an AI-powered teacher support tool. The theoretical foundation draws on John Biggs' constructive alignment framework, the revised Bloom's taxonomy, the AIAS scale, and the AAAE framework. It is argued that an assignment's vulnerability to AI delegation is a didactic rather than a technical problem. A typology of learning assignment transformation is systematized, comprising five types: "localization and personal experience," "critical analysis of an AI-generated response," "visibility of process," "multiformat presentation," and "oral defense." The practical embodiment of the typology is a specialized AI tool (Gem-bot) built on the Google Gemini platform, which implements three sequential operations: diagnosing assignment vulnerability, generating five transformation variants, and formulating assessment criteria for each, including quality performance markers and indicators of probable AI delegation. A key feature of the tool is that it does not replace the instructor in decision-making but structures the process, leaving the final choice to the professional. The proposed approach is oriented toward practical use by higher education instructors and can be adapted to various disciplines and student preparation levels.