(C) South China Morning Post

DAEGU, South Korea — On university campuses, the use of generative AI such as ChatGPT and Gemini has transitioned from a novel trend to an everyday essential. As the academic world wraps up the 2025 winter semester, a new dilemma is surfacing for faculty: how to balance the undeniable utility of AI with the preservation of critical thinking and academic integrity.

The "New Normal" on Campus
"To be honest, it feels like 10 out of 10 students use AI for their assignments," says Student A, a senior at Kyungpook National University. "I use both ChatGPT and Gemini to cross-verify data. It’s been the standard since 2023."

This sentiment is echoed across regional universities. Students report using AI for everything from brainstorming exam questions to drafting entire reports. In response, some professors have adopted a "if you can't beat them, join them" approach. A junior at Yeungnam University noted that younger faculty members now require students to submit their full AI chat logs alongside their essays. This allows instructors to verify whether the AI was used as a research assistant or if the student simply "copy-pasted" the output.

The Struggle for Evaluation and Ethics
Despite these creative solutions, the academic community faces significant hurdles. While AI can streamline research, there are growing concerns over "AI dependency," which may lead to a decline in analytical skills.

"In the corporate world, AI literacy is now a requirement, so banning it would be counter-intuitive," explained an Economics professor in Daegu. "However, in large lecture halls, it is difficult to implement alternative assessments like team debates or oral presentations to verify a student's true understanding."

Furthermore, the "AI arms race" is complicating grading. Professors admit that while basic AI patterns are easy to spot, high-end paid models produce sophisticated work that is nearly indistinguishable from human writing, creating an uneven playing field based on a student’s ability to pay for premium services.

The Path Forward: Hybrid Assessment
Global education experts suggest that the solution lies in a "Hybrid Evaluation" model. This involves bifurcating assessments into two categories:

AI-Excluded Proctored Exams: Traditional in-person tests to measure fundamental knowledge and raw reasoning.
AI-Integrated Projects: Assignments where students are judged on their ability to prompt, critique, and refine AI-generated content.

Kim Myung-joo, Director of the AI Safety Institute, emphasizes the importance of "Critical AI Literacy." He argues that the focus of modern education should shift toward teaching students how to skeptically review and edit AI outputs rather than accepting them as absolute truths.

As AI continues to evolve, South Korean universities are at a crossroads, tasked with redefining what it means to be "intelligent" in an age where answers are only a prompt away.

[Copyright (c) Global Economic Times. All Rights Reserved.]