Vietnam to pilot AI education across grades 1 to 12 from December 2025

AI learning reaches the full school pathway

Vietnam will begin piloting artificial intelligence education in selected schools in December 2025, covering grades 1 through 12. The trial will run until May 2026, followed by a Ministry of Education and Training review in June 2026 to refine the framework and prepare recommendations for expansion in future school years.

The ministry is linking the pilot to a broader digital transformation drive aimed at building digital literacy and modern skills at scale. Resolution 71, issued by the Politburo, sets a direction for stronger use of digital technology and AI in education management across the country. The school pilot is a central part of that agenda.

Students will encounter AI through existing subjects, projects and clubs, rather than a separate, demanding new subject. The ministry says the content should match current learning outcomes and should not overload school timetables. Schools will have flexibility to choose activities that fit their teaching conditions and resources.

AI already appears in everyday tools, including translation services, image recognition and systems that recommend content. The pilot aims to help children understand the ideas behind those tools, use them safely and responsibly, and gain confidence in learning. For younger pupils, that means activities connected to familiar classroom experiences, with attention to privacy, fairness and the role people play in guiding technology.

What happens between December and June?

The pilot spans the full general education pathway, divided into basic education and career orientation. Primary school and lower secondary school form the basic education stage. Upper secondary school forms the career orientation stage, where students will explore more advanced problems and possible links to future work.

The timetable sets out three steps:

• December 2025: AI education begins in selected schools across grades 1 through 12.
• December 2025 to May 2026: schools test approaches through computer science, STEM, elective topics, projects and extracurricular clubs.
• June 2026: the ministry reviews results, refines the framework and prepares recommendations for wider use in future school years.

The number and names of participating schools, and a date for wider rollout, have not been specified. The review is intended to help shape what comes after the trial, using lessons from how schools deliver the content.

Four strands, with different goals by age

The framework has four strands running through all grades: thinking that puts people at the center, AI ethics, AI techniques and applications, and AI system design. Together, they present AI as tools created and guided by people, with both benefits and limits. Students will also practice understanding data, solving problems and working in teams.

Primary pupils will meet AI through simple examples such as image and voice recognition. Early exposure does not mean asking children in grade 1 to undertake advanced coding. Lessons will help them understand that humans create AI and that personal information needs protection.

Students in lower secondary school, covering grades 6 to 9, will begin working with data and algorithms. An algorithm is a set of instructions used to carry out a task. Students will try AI tools to support assignments while learning to recognize risks, mistakes and bias in the results.

Upper secondary students will move toward designing simple AI systems, from models that follow rules to basic machine learning projects. A model that follows rules acts on instructions set by people, while machine learning uses patterns in data. Students will practice more advanced problem solving and begin mapping possible careers in technology and related fields.

The progression connects practical use with judgment. A child recognizing an AI feature, a teenager checking an answer and an older student designing a simple system will be working at different levels within the same four strands.

How AI fits into existing lessons

Schools are encouraged to connect AI activities to goals already in the curriculum. A science teacher might use a visual AI tool to classify images during a biology lesson. A mathematics teacher could explore pattern recognition by comparing a human strategy with the approach taken by a simple model.

A literature teacher might use a tool that converts speech into text to support oral presentations and feedback. Projects and themed weeks could bring several subjects together around practical tasks, such as creating a simple traffic sign detector or developing a classroom policy on privacy.

The ministry advises localities to make the most of existing resources and avoid scattered investment. Schools can use current laboratories and devices, adopt open learning materials, and work with partners on training and practical activities. The aim is to support learning without creating extra pressure for students or teachers.

Flexibility extends beyond ordinary lessons. Schools can integrate individual topics into subjects, organize AI clubs or hold community workshops. Each school can choose methods suited to its circumstances while keeping activities connected to existing learning outcomes.

Teachers face new tools and assessment choices

Teachers will be central to the pilot. The ministry plans to train core teaching teams and test AI tools that can help with lesson planning, classroom management and new forms of assessment. Those tasks bring AI into the work of educators as well as the work students complete.

Generative AI, which produces content in response to instructions, already presents teachers with questions about student assignments. A practical approach described in the plan is to reward original thinking while allowing guided AI use with clear disclosure of how a tool was used.

Vietnamese higher education has tested structured approaches to generative AI in assessment. One model sets levels ranging from no AI use to full AI use, with tasks designed to keep people involved in directing and checking the work. That research concerns universities, but its principles can also inform school assignments.

Teachers can ask for a personal voice, data collected in the real world or an oral defense in which students explain their work. AI might be permitted for brainstorming or formatting, with students required to explain what they used and why. Simple marking criteria that include a disclosure step can help reduce misconduct while developing digital judgment.

Privacy, bias and human oversight

Students and teachers will need to understand privacy, consent and bias alongside the benefits of AI. Personal data can include names, photographs, voice recordings, test scores and location. An AI tutor may record a student's interactions and performance, making the handling of that information part of the school's responsibility.

Schools should choose tools with clear data policies, avoid collecting sensitive information without a sound reason and prefer options that allow local control of information. Those choices matter when classroom activities involve student work or personal details.

Bias can appear when a model produces different results for different groups or repeats stereotypes. Lessons can help students ask who built a model, what information it learned from and where it might fail. Checking those points gives students a practical way to examine an AI answer instead of accepting it automatically.

A school digital code of conduct can set straightforward expectations for disclosure, citation and respectful use. Teacher oversight remains vital. The ministry also plans to test AI in school management, where safeguards will be needed so that automation supports human judgment.

Making access work beyond major cities

Vietnam aims to make AI learning available to all students, including those in rural and mountainous areas. Delivering activities under different local conditions is therefore part of the pilot's focus on flexibility and the use of facilities schools already have.

Recent projects involving Vietnamese and UK researchers have worked with schools across regions to develop teaching practices for AI tools together. Their focus includes teacher capacity, gender inclusion and practical resources suited to different settings. That approach aligns with guidance allowing schools to choose delivery models that fit their circumstances.

Schools can rotate devices, allocate laboratory time fairly and use content that works offline. Such arrangements provide practical ways to share equipment and plan activities around the resources available.

Local telecommunications partners can help with inexpensive connectivity plans for education platforms. Vietnamese technology companies can support schools through free access tiers and teacher training. Learning materials should be available in Vietnamese, with options to support ethnic minority languages, so that language and location are considered alongside equipment.

These measures are intended to let schools in cities, smaller towns and more remote areas participate with confidence. Teacher preparation, suitable materials and access to devices all form part of making the classroom activities workable.

What comparisons with China and the US offer

China and the United States provide different reference points for Vietnam. China integrates AI content into existing subjects, builds national platforms and coordinates policy centrally. The United States takes a decentralized approach, relying on decisions by states and school districts, with substantial participation from private companies.

Vietnam can draw on elements that fit its own school system. Integrating AI into current subjects can keep the program manageable, while a national competency framework can define what students should know at each level.

Partnerships between public bodies and private companies can bring equipment, platforms and teacher training. The state would set standards for safety and quality. Those roles connect the pilot's focus on shared resources with the need for consistent expectations about what schools teach and how they use technology.

Ho Chi Minh City's practical examples

Ho Chi Minh City already offers examples of how digital tools and AI can support school services. The city has used online enrollment across districts, with chat systems that answer questions from parents and provide guidance.

A map based system helps families find nearby schools and supports planning. These services connect technology to routine decisions families and education officials need to make, including where children can attend school.

The city's education department is also testing two AI solutions related to learning. One examines how students interact with educational content and suggests materials and schedules tailored to each learner. The other analyzes historical data to identify topics students find difficult, helping teachers plan targeted lessons.

The two trials address different parts of teaching: choosing resources and schedules for individual learners, and identifying subject areas that need more attention. Together with enrollment services, they illustrate practical uses of AI in both administration and instruction at scale.

What the June review can measure

The ministry plans to review the pilot after it ends in May 2026, with the review and framework refinement scheduled for June. Evidence from schools will be central to assessing whether the activities build skills while fitting within existing teaching time.

Useful measures include student skills by grade band, teacher confidence and completion of training. Another measure is the share of schools able to deliver activities with their available devices. Schools can also track how many lessons integrate AI without reducing time for core subjects.

Sample student work, short surveys and focus groups can help show how the pilot is working in practice. Information collected for that purpose should be anonymized, meaning identifying details are removed, and handled with care.

The review will help the ministry refine the framework and prepare recommendations for future expansion. It also provides a point to examine whether flexible delivery, teacher training and existing resources have been enough to support the activities tested.

How families can support the learning

Parents can support the pilot by discussing privacy and fairness with their children and encouraging independent thinking. Asking a child to explain how an AI tool arrived at a suggestion can make those ideas part of an everyday conversation.

A child might compare answers from two tools or check an AI summary against the original text. These simple activities create opportunities to notice differences, identify mistakes and explain why an answer is useful or needs checking.

Families can also learn their school's rules on disclosure and help students follow them. Knowing when AI use is permitted, and how students should describe that use, connects classroom expectations with work completed at home.

For schools and families alike, the pilot combines learning about AI with learning how to question it. Privacy, fairness, independent thought and teacher guidance remain part of the program from the earliest grades through preparation for future careers.

What to Know

• Vietnam's pilot will run from December 2025 to May 2026 in selected schools across grades 1 through 12.
• The Ministry of Education and Training plans a review in June 2026 to refine the framework and guide recommendations for expansion.
• Four strands cover thinking that puts people at the center, AI ethics, techniques and applications, and system design.
• Learning will progress from simple recognition tools and privacy awareness to data, algorithms and basic AI projects.
• Schools will integrate AI into existing lessons, projects, elective activities and clubs without overloading timetables.
• Teachers will receive training, while AI tools will be tested for planning, classroom management and assessment.
• Privacy, bias, disclosure and human oversight are central concerns.
• Flexible resources and teaching approaches aim to include students in urban, rural and mountainous areas.