'AI abuse can artificially increase the number of good and excellent degrees'

lạm dụng AI - Ảnh 1. — Prof. Dr. Bang Tien Long speaks at the conference - Photo: NGUYEN BAO

On October 16, the Center for Education Quality Assessment - Association of Vietnamese Universities and Colleges held a ceremony to celebrate its 10th anniversary and a workshop on "Trends in education quality assessment for the period 2026-2030 with the help of artificial intelligence".

When machines answer better than students

Sharing at the conference, Professor Dang Ung Van told the story of "when machines answered better than students" in a class in Central Vietnam.

When a philosophy lecturer assigned students to write an essay on authenticity in the digital age, one student submitted an essay with "unusual" fluency, with a rich vocabulary, a coherent structure, and many references that very few Vietnamese students know. When asked by the lecturer, the student said he had asked AI to do it, only editing a few sentences about his writing style.

From the above example, Mr. Van thinks that the problem is quite difficult, should the lecturer fail the student for using Chat GPT, the purpose of education is to demonstrate originality, thinking ability or meaningful participation in ideas, even ideas suggested by the machine?

According to Mr. Van, liberal education emphasizes the comprehensive development of human beings (reason, emotion, morality, freedom of thought), but AI can perform complex tasks such as: writing, analyzing, solving problems. Students can be tempted to do more than intellectual work by asking AI to do it for them.

“The emergence of AI is not only a technological advancement but also a philosophical challenge for higher education. AI will not replace education, but will reshape the reasons and methods of education. This requires a new philosophy – not replacing the old, but connecting, interpreting and expanding them in new contexts,” said Mr. Van.

Similarly, Professor Dr. Bang Tien Long - former permanent deputy minister of the Ministry of Education and Training, said AI also brings with it challenges regarding academic integrity.

"If not managed, the misuse of AI can lead to the risk of grade inflation, reducing the real value of university degrees, when the rate of excellent degrees increases artificially," said Mr. Long.

Many opportunities when applying AI to university accreditation

Regarding the accreditation of training programs/university education institutions, Mr. Bang Tien Long commented that the traditional accreditation model has many limitations, including wasting time and human resources.

In addition, the 5-10 year evaluation cycle creates a monitoring gap; lacks flexibility, does not reflect continuous quality; and makes it difficult to respond quickly to social changes.

Therefore, Mr. Long believes that it is necessary to shift from a static model to a dynamic model in quality assurance, and it is necessary to bring AI into educational quality assessment.

"AI can process large volumes of reports, detect abnormalities during training; continuously monitor, track indicators in real time and give early warnings, instead of relying on the current 5-10 year cycle," said Mr. Long.

According to Mr. Long, the implementation of AI in higher education requires a strategy for each group, in which, for state management agencies, it is necessary to complete the legal framework, specific regulations on the use of AI; and ethical guidelines for the application of AI in education.

For universities, it is necessary to develop internal policies, guide students and lecturers on how to use technology responsibly. At the same time, integrate ethical rules when using AI into the school's academic regulations.

In addition, universities need to invest in data infrastructure and technology to comprehensively digitalize processes; organize regular AI training courses for lecturers.

For lecturers and students, it is necessary to innovate methods of assessing and developing AI capacity in teaching and learning. Consider AI as a tool to support work and study but not rely on it.

Many innovations in education quality assessment

Speaking at the workshop, Prof. Dr. Huynh Van Chuong - Director of the Department of Quality Management ( Ministry of Education and Training ) said that it is expected that from 2026, the assessment of training programs and higher education institutions will have many major adjustments.

In which, each evaluation criterion has only two levels "pass" or "fail". Previously, applying 7 levels was controversial and difficult to standardize.

Bringing the level to two, pass or fail, will simplify guidance, make results transparent, and be consistent with international practices (US and Europe).

Besides, setting up "conditional criteria" will help focus on core factors such as output standards,... If the conditional criteria are not met, the standard will not be met.

Regarding the classification of training programs/educational institutions after accreditation, there will be three levels of results, including: passed, conditionally passed, and failed. Schools have 24 months to improve if they are ranked "conditionally passed".

Back to topic

NGUYEN BAO

Source: https://tuoitre.vn/lam-dung-ai-co-the-khien-bang-gioi-xuat-sac-tang-len-mot-cach-gia-tao-20251016171120736.htm