By abusing AI, many young people graduate bitter because of their empty knowledge

(Dan Tri) - Faced with the uncontrolled use of AI by students and even lecturers, the regulations effective from August 15 set a series of new requirements on transparency and integrity when using technology.

Beautiful degree but empty knowledge

As a student who has just graduated from a university in Hanoi, P.U. (22 years old) said she started using AI in the middle of her second year.

Initially, this tool was only used to find documentation or implementation suggestions. However, as time goes on, AI gradually takes over almost the entire test-taking process.

"AI participates in up to 80% of assignments. Group assignments, essays or even dissertations, I almost always use AI. Some assignments just edit the words and then submit them," U. shared.

To avoid being discovered by the lecturer, she often edits her writing style, adds a few personal opinions, cuts out sentences that are too elaborate, and even intentionally makes spelling mistakes.

She admitted that even though she relies on AI, it saves a lot of time and still scores enough to make herself satisfied.

Many times lecturers asked questions when they suspected that U.'s work was supported by machines. However, she insisted on denying it and claiming that it was her own effort.

Just like that, from subject to subject, she used AI with more frequency and affirmed that she could not return to the self-study and self-research method like the first year.

Only when she graduated, held a good degree in her hand and started working, did U. realize that she had a lot of specialized knowledge that she almost didn't know.

When her manager assigned her a job, she didn't know where to start.

"When I was in school, I only thought about how to complete my assignments on time and keep my grades up. When I graduated, I saw that the degree was just one part, but in terms of knowledge, it was completely empty," said P.U. confide.

Even though he has found a job, U. is still struggling to handle requests that require knowledge that he previously learned but no longer retains.

The fact that students are increasingly attached to AI in learning and research does not only appear in individual stories.

Research published in 2026 on Data in Brief, surveying 1,580 Vietnamese students, showed that more than 83% of participants had used AI for at least a year.

From the above numbers, the authors raised concerns about the risk of students using AI passively, instead of searching, synthesizing and processing information themselves, thereby weakening independent thinking.

This concern was partly expressed in an experiment published in 2025 in Social Sciences & Humanities Open.

After 45 days, the group of students using ChatGPT only answered an average of 57.5% of the questions correctly, lower than the 68.5% of the group studying using the traditional method.

The research team suggests that when AI takes over part of the thinking process, learners may bypass the cognitive effort required to form and retain expert knowledge.

The authors also introduce the concept of "borrowing capacity". The term refers to the fact that AI provides structure and coherent expression, helping learners create products that seem complete and have a sense of mastery of knowledge, while the actual level of understanding and memory is not adequate.

In P.U.'s case, the weakness caused by using AI only really becomes apparent when actual job requirements appear.

The degree and grades are still there, but professional knowledge and practical experience are equally important.

A study published in 2026 in the North American Journal of Psychology, surveying 134 Vietnamese employers, also showed that the majority of graduates' abilities were only assessed at an average level.

The areas that are underrated are conflict handling and especially the ability to transition from the university environment to real work.

Even instructors can abuse AI

Besides learners, AI abuse and fraud are also starting to be noticed by some students and lecturers.

H.M., a student who just completed his thesis and graduated from a university in Hanoi, said he repeatedly received feedback from his instructor with a writing style similar to ChatGPT.

"One time, the teacher sent the original comment containing the sentence "This edit is more reasonable than your old outline". I was quite surprised because this is the sentence AI often responds to when asked for feedback," H.M. tell.

She commented that this content was like a general suggestion, after reading it she still didn't know what to do next.

In particular, she said that at the first meeting, the lecturers agreed on the implementation plan.

However, a few days later the lecturer sent a copy of instructions completely different from what had been finalized and added a series of requirements that were beyond our capabilities.

H.M. believes that this sudden change makes students not know which direction to follow.

When she asked again, she only received the response that "I can completely deploy it with the support of AI".

From August 15, not disclosing the use of AI may be punished

Faced with the uncontrolled use of AI, from August 15, the obligation to be transparent when using this technology will be set out more clearly for both learners and lecturers.

According to many education experts, AI is still a useful tool if used to suggest ideas, check for errors or support research. It is important that learners still have to analyze, evaluate and take responsibility for the final product.

According to Circular No. 49/2026/TT-BGDDT, the act of not disclosing or incompletely disclosing the use of technology and artificial intelligence when required according to internal regulations is a violation of academic integrity.

This regulation does not mean that students and lecturers must declare their use of AI in all cases.

Accordingly, the obligation to publish is imposed when the school, course, lecturer or internal regulations have specific requirements.

However, once a claim is made, the parties must be honest about the extent to which technology is involved in the product.

At the same time, the Circular emphasizes that violations of academic integrity in the digital environment also include:

- Using technology, including artificial intelligence, to cheat in studies, tests, assessments or scientific research.

- Copying, plagiarism, unauthorized use of documents, data, digital learning materials or research results.

- Forging or falsifying data, research results or learning information.

- Ask someone else to perform or perform on your behalf learning, testing, assessment or research tasks.

- Other acts that violate regulations on academic integrity according to the provisions of law and regulations of educational institutions.

However, the Circular does not provide a general level of discipline applicable to all educational establishments. Instead, each school must issue internal regulations on academic integrity, clearly defining violations, prevention methods, handling procedures and responsibilities of the parties.

Schools must also develop regulations or codes of conduct on using AI in training, testing, assessment, scientific research, administration and service provision.

The circular also requires schools to promote the development of digital and AI capabilities for learners, lecturers, and administrators.