"Teach us how to use AI" — this request is rapidly growing at universities and TVET (Technical and Vocational Education and Training) institutions across Laos. Yet in most cases, the instructors themselves have no hands-on experience using AI in their own work.

Through our AI talent development program for businesses, we have put into practice methods for building AI literacy among non-engineers. What this process has revealed is that the underlying assumptions differ significantly between corporate training and educational institutions. Corporate training demands immediate results — skills employees can apply on the job tomorrow — whereas educational institutions aim to cultivate foundational capabilities that graduates can draw on for five years after leaving school.

This article draws on insights gained from corporate training settings to propose design principles for integrating AI literacy into curricula within the respective contexts of TVET institutions and universities.

When educational institutions in Laos confront AI, they must first accurately grasp the current situation.

The degree of AI education adoption in higher education within ASEAN varies significantly by country. Singapore, Malaysia, and Thailand have already established dedicated AI departments and made AI literacy courses compulsory for all faculties.

The situation in Laos is as follows:

• National University of Laos (NUOL): The Faculty of Information Technology offers programming courses, but specialized subjects in AI and machine learning are limited.
• TVET institutions: ICT-related curricula exist, but the content centers primarily on PC operation and Office software.
• Private universities: Some data science courses are offered, but the materials are in English, creating a high access barrier for students.

This gap is both a threat and an opportunity. By drawing on AI education best practices that other countries have built up over several years, Laos can take a "shortcut" tailored to its own context.

AI literacy is not about "being able to program." It is the ability to understand the basic mechanisms of AI, appropriately use AI tools in one's work and daily life, and critically evaluate AI outputs.

Considering Laos's economic structure (agriculture 15%, services 42%, industry 43%), AI literacy is not only needed by engineers.

• Agricultural workers: Reading satellite data, utilizing demand forecasting (→ Agriculture × AI)
• Financial operations staff: Document processing via AI-OCR, fraud detection mechanisms (→ Financial DX)
• Tourism and hospitality staff: Operating AI chatbots, designing multilingual support (→ Tourism DX)
• Accounting staff: Implementing accounting automation, reviewing AI outputs (→ Accounting AI Automation)

The role of educational institutions is to produce these "AI-capable individuals" across all occupations.

When applying corporate AI training methods to educational institutions, three key differences must be considered.

In corporate training, a cycle of "apply today what you learned today" is in motion, whereas in educational institutions there is a time lag where "what you learned only becomes useful after graduation." It is internships and industry-academia collaborative projects that bridge this gap.

In corporate training, evaluation is based on "whether it can be applied on the job" rather than "test scores." Educational institutions, on the other hand, require formal grade assessment.

For evaluating AI literacy, the following is recommended instead of traditional written exams:

• Project-based assessment (40%): Conduct team-based problem-solving projects using AI, evaluated on deliverables and presentations
• Portfolio assessment (30%): Accumulate a record of AI-assisted work throughout the semester, evaluated on the trajectory of growth
• Knowledge test (20%): A comprehension test covering basic AI concepts, ethics, and limitations
• Peer review (10%): Evaluation of contributions among team members

This assessment method allows for a distinction between "having knowledge of AI" and "being able to use AI."

Among educational institutions, in settings like Technical and Vocational Education and Training (TVET) schools where the curriculum is directly linked to practical work, "skills applicable in the workplace" take priority over "academic understanding." The hands-on, practice-oriented approach that has proven effective in corporate training can also be applied to curriculum design in these educational institutions.

TVET's AI literacy curriculum is directly linked to job-specific work scenarios.

Manufacturing Course (4 months):

Tourism Course (3 months):

The key point is that coding is not taught. What TVET students need is the ability to effectively use existing AI tools in their work — not the ability to develop AI.

Many TVET teachers have no personal experience using AI. This is not a challenge unique to Laos—it is common across ASEAN as a whole.

Training Design for Teachers (Develop Teachers First):

1. Phase 1 (2 weeks): Teachers themselves use AI chatbots extensively for daily tasks (lesson planning, grade management, parent communication)
2. Phase 2 (2 weeks): Workshops on instructional design using AI. Share case studies with other teachers
3. Phase 3 (ongoing): Monthly teacher community sessions to share successful and unsuccessful cases

In corporate training settings, it has been demonstrated that even non-engineers can develop lasting AI literacy. Teachers likewise do not need to become "AI experts"—they simply need to become "facilitators who incorporate AI into their lessons."

At the university level, a deeper level of understanding and applied skills are developed compared to TVET.

AI literacy must not remain a "specialized subject" confined to the Faculty of Information Engineering. The National University of Laos (NUOL) has faculties of Economics, Law, Agriculture, and Medicine. AI is utilized in different ways across each of these faculties.

Sample Cross-Faculty AI Course Design (1 Semester · 15 Sessions):

The advantage of cross-faculty mixed teams lies in the intersection of technical perspectives (Information Engineering) and domain knowledge (Economics, Law, Agriculture). For example, in a project on "Applying AI to Agriculture in Laos," students from the Faculty of Agriculture define the challenges, students from the Faculty of Information Engineering propose technical implementation approaches, and students from the Faculty of Law review the legal requirements for data protection.

The most effective approach in university AI education is to have students tackle real corporate challenges as PoCs.

Industry-Academia PoC Operating Model:

1. Companies present their challenges — For example: "We want to improve warehouse inventory management efficiency" or "We want to automatically classify customer inquiries"
2. Faculty design the PoC scope — Narrowing the scope to what can be completed within a single semester
3. Student teams (3–5 members) execute the PoC — Data collection → Analysis → Prototype development → Results presentation
4. Companies provide feedback — Evaluation from a practical perspective, with internship opportunities offered to outstanding teams

The "how to conduct proof of concept" approach explained in the PoC development article functions directly as an educational program. It is a win-win structure that provides students with practical experience while giving companies an opportunity to validate their AI applications.

One of the biggest hurdles in AI education in Laos is the shortage of educational materials in the Lao language.

Creating Lao-language AI educational materials from scratch is costly. A practical approach is to localize existing high-quality English materials into Lao.

Recommended material sources:

• Google AI Education — Explains AI fundamentals for non-engineers. Available for translation and adaptation under a CC BY license
• Elements of AI (University of Helsinki) — An introductory AI course for non-technical audiences. Already translated into multiple languages
• fast.ai — The Practical Deep Learning course. Hands-on focused and ideal for teachers' self-study

During localization, draw on the multilingual expertise developed in Building a Lao Language AI Chatbot. By using AI to translate English materials into Lao and establishing a workflow where teachers review and revise the output, the pace of material development can be significantly accelerated.

It is also important to incorporate Lao-specific examples (agriculture, finance, tourism). Replacing Silicon Valley case studies from English materials with Lao contexts will substantially improve students' comprehension and engagement.

Not just "understanding how AI works" but hands-on experience of "actually trying AI" is essential in education. A demo environment operable in Lao will be prepared.

Minimal AI Demo Environment:

• AI Chatbot Experience: Have students use the free versions of Claude or ChatGPT in Lao. Simply teaching them how to write prompts is enough for students to begin exploring autonomously.
• Image Recognition Experience: Have students build their own object recognition models using a camera with Google Teachable Machine (browser-based, free). No coding required.
• Data Analysis Experience: Visualize publicly available Lao data (weather data, agricultural product prices) using Google Sheets + Google Colaboratory.

For TVET schools in rural areas with unstable internet connections, it is also effective to pre-install lightweight models that run offline on classroom PCs. There is also the option of sharing AWS instances introduced in the Cloud Migration Guide among educational institutions.

Here are two recurring failure patterns seen in the introduction of AI education.

Too much time is spent teaching AI theory (the mathematical foundations of neural networks, classifications of machine learning algorithms), and the semester ends without students ever experiencing firsthand what AI can actually do.

What I recommend for educational institutions in Laos is a "7:3 rule" — dedicating 70% of class time to hands-on work (exercises using actual AI tools) and 30% to theory. Theory should be positioned before and after the hands-on portions, serving as an explanation of "why things work the way they do."

Insights from corporate AI training also show that lecture-centered programs tend to have low retention rates, with a marked improvement in retention when programs shift to a hands-on-centered approach. Similar effects can be expected in educational institutions as well.

Importing AI education curricula from developed countries as-is will not work in Laos. For example:

• Exercises using large datasets — The classroom Wi-Fi is slow, making downloads take an hour
• Model training using GPUs — The lab PCs are not equipped with GPUs
• Real-time processing using cloud APIs — Power outages cut off internet access

When designing curricula, incorporate Laos's IT environment constraints as preconditions. By centering the design around offline-capable exercises, lightweight models, and mobile device operation, infrastructure constraints need not become curriculum constraints.

Q1: Is prerequisite knowledge of programming required for AI literacy courses?

No. AI literacy education at the TVET level does not teach coding at all. The focus is on "effectively using" existing AI tools (chatbots, image recognition, translation tools). Even at the university level, cross-departmental courses center on no-code tools. Programming languages such as Python are limited to specialized courses in the Faculty of Information Engineering.

Q2: If teachers lack AI skills, are external instructors necessary?

While support from external instructors is effective during the initial introduction phase, the goal should be to reach a state where "teachers themselves can teach" from a sustainability perspective. The recommended approach is to conduct intensive training for teachers (2–4 weeks), allowing them to become comfortable using AI in their own work before moving on to course design.

Q3: Is it possible to secure a budget for AI education through JICA or ADB educational support programs?

ADB is implementing multiple support projects under its TVET reform program in Laos, and ICT curriculum strengthening is included among its priority areas. JICA also supports digital human resource development within its framework for higher education assistance. A key factor for proposal adoption is incorporating specific outcome indicators for AI literacy—such as the number of program completers, employment rates, and employer satisfaction—into the proposal.

Laos's AI education is not "behind" — it is at a stage where it can be "designed from scratch." There is an opportunity to learn from the trial and error of other countries and build a curriculum optimized for the Laotian context.

3 actions educational institutions can start today:

1. Have teachers start using AI — Begin by using AI chatbots for lesson planning and grade management, so that teachers themselves can experience the "value of AI" firsthand
2. Pilot just one subject — Rather than overhauling the entire curriculum, incorporate AI literacy elements into a single subject and measure the results
3. Leverage corporate AI training expertise — Partner with companies considering AI adoption and providers offering corporate AI training to obtain practical teaching materials and instructor support

The goal of AI education is not to "mass-produce AI experts." It is to develop people across all occupations who can "use AI as a tool." By combining the practical AI knowledge that companies have refined on the ground with the systematic curriculum design capabilities of educational institutions, it is possible to create a realistic impact on human resource development in Laos.