Abstract:

The rapid expansion of Artificial Intelligence (AI)-based tools is significantly reshaping teaching and assessment processes in higher education. Previous research conducted by the authors, presented at EDULEARN 2025, showed that the widespread availability of these technologies drastically reduces the time required to complete traditional academic tasks in technical disciplines, thereby challenging the validity of conventional assessment models in technological postgraduate programs.

Building on these findings, the present study examines the redesign and implementation of new assessment strategies applied during recent academic cycles in postgraduate programs related to Cybersecurity, Technological Project Management, and Information Technologies in institutions in Argentina and Paraguay.

The study analyses how traditional evaluations centred on the individual completion of technical assignments have been progressively replaced by assessment models oriented toward complex professional contexts. Implemented strategies include scenario-based evaluations, iterative challenges incorporating dynamic variables, practical simulations focused on decision-making, and collaborative activities requiring critical analysis and technical argumentation. These approaches were designed to move assessment away from the verification of routine technical outputs and toward the appraisal of higher-order professional competencies that are directly relevant to contemporary technological practice and to AI-mediated professional environments.

The results indicate that this redesign enables the assessment of competencies that cannot be fully automated by AI-based tools, including the strategic interpretation of technological problems, the justification of technical decisions, the management of uncertainty, and the critical integration of automated technologies within professional environments.

Based on the pedagogical experience derived from these redesign processes, the study introduces the concept of an AI-Compatible Assessment Protocol (AICAP), conceived as a conceptual framework for designing assessment models suitable for educational environments where artificial intelligence is integrated into the learning process. The protocol seeks to facilitate interoperability between assessment approaches implemented across different academic institutions, contributing to the development of standards adapted to highly digitalised higher education systems.

The proposed AICAP framework is structured around four principles:
a. professional contextualisation of assessment,
b. evaluation of reasoning and decision-making processes,
c. critical integration of artificial intelligence, and
d. institutional adaptability and interoperability, enabling assessment models to be transferable across programs and educational contexts.

This research contributes to the academic discussion on how higher education assessment models must evolve in response to AI-driven technological disruption, presenting empirical insights and proposing an initial framework for AI-compatible assessment in technological postgraduate education.

Keywords:

Artificial Intelligence, Assessment, Competencies, Interoperability, Postgraduate.