Is ChatGPT-4 Capable of Passing the Entrance Exams for the Public Teaching Career (2018, 2019, 2022, and 2024)?

Abstract

This article examines the performance of ChatGPT-4 in the entrance examinations for the public teaching career in Peru (2018, 2019, 2022, and 2024) using a descriptive, cross-sectional approach. Responses in Mathematics, Science and Technology (CTA), and Communication were evaluated and compared against official scoring standards. The results indicate that ChatGPT-4 passed all examinations, achieving its highest scores in Science and Technology (184 points in 2024) and in the Primary level (173.5 points in 2018). However, lower performance was observed in Mathematics (137 points in 2022 and 150 points in 2024) and Communication (155 points in 2024), suggesting limitations in critical analysis and contextual interpretation. The findings highlight the need to train language models in more complex and context-sensitive scenarios, as well as to reconsider the design of assessment instruments. Overall, ChatGPT-4 demonstrates potential as an educational support tool, provided its use is guided by ethical principles.