Vietnam universities need to teach students responsible AI use

The circular, issued on June 30, 2026, sets standards and procedures for higher education training programs at the bachelor's, master's and doctoral levels, covering their development, approval, implementation, and oversight.

It requires universities to include digital competency, AI-related content, foreign-language proficiency, adaptability, and lifelong learning in the learning outcomes for undergraduate programs.

As an English lecturer specializing in teaching methodology, I believe this is a necessary and timely policy.

But the success of AI integration should not be measured by how many tools are introduced into classrooms.

What matters is whether students are prepared to use AI in an informed, critical, creative, and responsible way.

AI has already entered university classrooms, even before many training programs formally addressed it.

Students use AI to translate texts, explain concepts, generate ideas, outline assignments, edit writing, and design presentations.

For students studying to become English language teachers, AI can also be used to prepare lesson plans, create communicative activities, design rubrics, develop teaching materials, and provide language feedback.

The key question, therefore, is no longer whether students should be allowed to use AI, but how universities can teach them to use the technology.

If AI integration is limited to a digital-skills course or a few sessions on writing prompts, the new learning outcomes could easily become a formal requirement rather than a meaningful change.

Students should become co-creators with AI

AI tools may change within months, but skills such as critical thinking, professional judgment, ethical responsibility, and self-learning will remain valuable.

Universities should not simply train students to become technology consumers who depend on ready-made answers.

They should prepare students to become responsible AI users and co-creators.

UNESCO's AI competency framework for students, published in 2024, identifies 12 competencies across four areas: a human-centered mindset, AI ethics, AI techniques and applications, and AI system design.

The competencies are developed at three levels: understand, apply, and create.

The framework aims to help students operate AI tools while also developing a sense of responsibility, foundational knowledge, and the ability to contribute to safe and sustainable AI solutions.

For students studying to become English language teachers, knowing how to ask AI to generate a lesson plan is only the starting point. 

They also need to assess whether the plan suits learners' proficiency levels, needs, cultural backgrounds, and learning objectives.

They should be able to identify inaccurate information, unsupported assumptions, and activities that may not work in a real classroom.

At a higher level, students should be able to revise AI-generated materials, combine them with their professional knowledge, and explain their pedagogical decisions.

For example, in a teaching methodology course, instead of simply asking students to submit a lesson plan, lecturers could ask them to create one with AI assistance and then analyze its strengths and weaknesses before revising it based on learning theories and the characteristics of a specific group of learners.

The final submission could include the lesson plan, an AI-use log, verified information sources, and an explanation of the pedagogical decisions.

In this way, AI becomes an object of critical thinking and a partner in the creative process, rather than a machine that does students' work for them.

The implementation of Circular 54/2026 should therefore begin at the curriculum level.

Each program should clearly define what graduates are expected to be able to do with digital technology and AI in their professional fields.

Generic statements such as 'be able to use AI' are difficult to observe and assess.

Learning outcomes should instead describe specific abilities, such as selecting appropriate tools, writing effective prompts, verifying information, identifying bias, protecting data, disclosing AI use, evaluating AI-generated content, and using AI to solve professional problems.

Assessment also needs to change

AI competency should be integrated throughout the curriculum instead of being introduced in a single course.

At the initial stage, students can learn basic AI concepts, data safety, and ethical principles.

At the intermediate stage, they can apply AI to tasks in individual courses.

By the final stage, students should demonstrate that they can use AI independently, critically, and appropriately in their professions through internships, projects or portfolios.

Assessment methods also need to change.

If lecturers continue assigning tasks that AI can complete within seconds, banning students from using AI will not solve the problem.

Universities should put greater emphasis on assessing the learning process, portfolios, real-world assignments, personal reflection, and direct oral defenses.

Students can be allowed to use AI but should document their interactions with the technology, verify sources, identify where AI was used, and take responsibility for the final product.

The focus of assessment should shift from asking, "Did the student use AI?" to asking, "How did the student use AI, and how sound was their judgment?"

Universities also need to invest in lecturers, teaching materials, infrastructure and data protection.

Lecturers cannot be expected to integrate AI creatively if they receive only a single technical training session and are then left to work things out on their own.

Universities should establish communities of practice, banks of sample assignments, shared rubrics and mechanisms to recognize the time lecturers spend redesigning courses.

Most importantly, AI integration should not overshadow the role of people in education.

In language teaching, for example, encouragement, empathy, interaction, and an understanding of learners remain qualities that AI cannot easily replace.

Teachers need to know how to use AI to expand their professional capabilities, but they must also know when not to use it.

A good pedagogical decision is not necessarily the one that uses the most technology.

It is the one that best serves learning objectives and students' development.

Circular 54/2026 gives universities an opportunity to rethink education in the age of AI.

To make the policy meaningful, universities need to shift from adding tools to developing competencies, from controlling AI use to guiding responsible use, and from broad learning outcomes to observable and assessable evidence.

Students should not simply consume products generated by technology.

They should become people who can think critically, take responsibility, and co-create with AI to address practical challenges in their professions and society.

* This article is authored by Le Truong An, who wrote it in Vietnamese. Tuoi Tre News provides the English translation.