121
CHALLENGES AND DIFFICULTIES IN APPLYING AI IN VOCATIONAL
EDUCATION
Le Huu Phuoc
Ho Chi Minh City University of Technology, Vietnam National University HCMC
Email: huuphuoc@hcmut.edu.vn; Phone: 0974948497
ABSTRACT
This paper analyzes the challenges and difficulties in applying artificial intelligence (AI) in
vocational education, including issues related to infrastructure, skills of teachers and
students, implementation costs, as well as ethical and legal barriers. The research indicates
that although AI brings many potentials such as personalized learning and automated
assessment, effective implementation still faces many obstacles. The article proposes several
solutions to overcome these challenges, including investing in training, developing
supportive policies, and raising awareness about the benefits of AI.
Keywords: Artificial intelligence, vocational education, challenges, difficulties, solutions.
INTRODUCTION
In the era of the 4.0 industrial revolution, artificial intelligence (AI) has become one of the
pioneering technologies, profoundly changing how people approach education, labor, and
many other fields. Vocational education, with its important role in training high-quality
workforce, is not outside this trend. AI promises to bring major breakthroughs, from
personalizing learning pathways, automating management processes, to simulating reality to
enhance practical skills of students. However, alongside its enormous potential, the
application of AI in vocational education also faces many challenges and difficulties, from
technical and financial to ethical and social issues. This is a reality that countries, including
Vietnam, need to face and find solutions to maximize the benefits that this technology
brings.
Vocational education plays a core role in preparing human resources for the labor market,
especially in the context of globalization and the rapid development of modern industries.
According to the International Labor Organization (ILO) 2023 report, more than 50% of
future jobs will require technical and digital skills, which vocational education needs to
address. However, many current vocational education systems still rely on traditional
methods, lack innovation, and fail to keep pace with technological developments. This is
why AI becomes a potential solution, with its ability to analyze data, predict trends, and
optimize the learning process.
AI has been successfully applied in many fields such as healthcare, finance, and
manufacturing, but in education, especially vocational education, implementation is still in
its infancy. AI tools, such as machine learning systems, educational chatbots, and virtual
reality environments, can assist teachers and students in enhancing training effectiveness.
However, these benefits do not come easily. Many countries, including Vietnam, face
barriers such as lack of infrastructure, teacher and student skills, as well as high costs to
deploy and maintain technology. Additionally, ethical issues, such as data security and
privacy, also raise major questions that need to be addressed.
122
In Vietnam, vocational education is under great pressure from labor market demands,
especially in industries such as information technology, machinery manufacturing, and
services. The government has introduced many policies to promote innovation in education,
but the integration of AI still faces many obstacles. For example, many vocational training
facilities in remote and mountainous areas lack modern equipment and stable internet
connections, while schools in major cities face difficulties in retraining teachers and
convincing students to accept new technology. These challenges not only slow down the
modernization process of vocational education but also create inequality in access to
learning opportunities.
Research Objectives
This paper aims to analyze and evaluate the challenges and difficulties in applying AI in
vocational education, and propose feasible solutions to overcome these barriers.
Specifically, the research will focus on:
Identifying technical, financial, and human resource issues related to AI deployment.
Analyzing ethical, legal, and social barriers that vocational education institutions
face.
Comparing international experiences and drawing lessons applicable to the
Vietnamese context.
Proposing specific recommendations for government, educational institutions, and
businesses to promote effective use of AI.
Research Scope
The research focuses on key aspects of AI application in vocational education, including
technological infrastructure, human capacity, implementation costs, and ethical issues. The
geographic scope is primarily Vietnam, with international comparisons from developed
countries like the US, Germany, and developing countries with similar contexts. The
research period is limited to the current period (2023-2024), based on the latest documents,
reports, and technology trends.
Research Significance
Studying the challenges and difficulties in applying AI not only helps policy makers and
education managers better understand the current situation but also provides a scientific
basis for building sustainable development strategies. In the context of Vietnam striving to
become a modern industrial country, integrating AI into vocational education is an important
step to improve human resource quality and compete internationally. However, if current
issues are not thoroughly addressed, the potential of AI may be wasted, or worse, cause
negative consequences such as increasing the digital divide and educational inequality.
CONTENT
1. The Potential of AI in Vocational Education
Artificial intelligence is becoming a revolutionary tool in many fields, including vocational
education - a field that plays an essential role in preparing the technical workforce for the
digital economy. According to a UNESCO report (2022), AI can help increase learning
efficiency by up to 30%, especially in training programs with high practical elements like
vocational training. In the context of the 4.0 industrial revolution, vocational education
123
needs to adapt quickly to provide modern skills that meet the constantly changing needs of
the labor market.
First, AI personalizes learning pathways. One of the prominent applications of AI is the
ability to design personalized learning programs. AI can analyze learner data such as
progress, assessment results, and learning habits to adjust content, pace, and teaching
methods. According to HolonIQ research (2021), more than 70% of vocational learning
platforms use AI to suggest courses and assess learners' skills in real-time. Systems like
Squirrel AI (China) have shown effectiveness by helping students increase their average
scores by 20-30% after 6 months of personalized learning.
Second, AI automates administrative tasks and assessments. AI helps save resources by
automating many administrative and assessment tasks. Systems like Gradescope (US) are
now widely used for automatic grading at more than 1,200 educational institutions.
McKinsey's report (2020) states that AI technologies can help reduce teachers'
administrative time by up to 40%, allowing them to focus on specialized teaching.
Third, AI simulates and provides practical reality. AI combined with simulation technologies
such as virtual reality (VR) and augmented reality (AR) helps students practice in a safe and
flexible environment. In Australia, TAFE Institute has used VR simulation models
integrated with AI to train refrigeration technicians, helping reduce training costs by 25%
and increase course completion rates by 18%. This technology is particularly effective in
high-risk industries such as aviation, mechanics, and healthcare.
Fourth, AI connects with the labor market. AI can analyze big data from the labor market
(job listings, industry forecasts) to adjust training programs. According to the World
Economic Forum (2023), AI helps predict skills that will be in high demand within the next
5 years, including data analysis, automated machine control, and cybersecurity. Platforms
like Burning Glass Technologies have applied AI to connect learners with suitable job
opportunities, contributing to a 20% reduction in unemployment rates in some pilot regions
in the US.
2. Challenges and Difficulties
Despite AI bringing many transformative opportunities for vocational education, deploying
this technology still faces a series of significant barriers across many aspects - from
technical infrastructure, human capacity to financial, legal, and social issues. These
challenges not only hinder the speed of AI application but also pose urgent requirements for
supportive policies and synchronous implementation directions at national and local levels.
2.1. Infrastructure and Technology
Chart 1: Percentage of vocational education institutions with suitable infrastructure
for AI deployment
124
One of the biggest barriers is that technical infrastructure conditions do not meet the
requirements for comprehensive AI deployment. According to a survey by Vietnam's
Ministry of Labor, War Invalids and Social Affairs (2022), only 17% of vocational
education institutions have infrastructure capable of deploying digital technology systems,
including big data storage systems, dedicated servers, high-speed internet connections, and
AI support software. This rate is significantly lower than the ASEAN regional average
(35%), showing that Vietnam is lagging in the process of digitizing vocational education.
In particular, the digital divide between regions exacerbates the problem. Up to 62% of
vocational training facilities in rural and mountainous areas do not have stable internet
connections or sufficient endpoint devices to deploy digital learning systems. This causes AI
- instead of being a tool to eliminate inequality - to risk increasing the educational access
gap between regions.
2.2. Skills and Capabilities of Teachers and Students
Chart 2: Percentage of teachers with digital technology skills
The lack of digital technology capacity of teachers and students is another serious barrier.
According to the Vietnam Institute of Educational Sciences (2021), only 28% of vocational
teachers are properly trained in digital technology, of which less than 10% have experience
with AI applications in teaching. The lack of foundational knowledge about AI not only
makes it difficult for teachers to integrate technology into lessons but also leads to anxiety
and hesitation in the digital transformation process.
On the student side, a survey by Microsoft Vietnam (2023) shows that only 12% of
vocational students feel confident using AI-based learning tools such as chatbots, adaptive
learning systems, or simulation software. Most students are still accustomed to traditional
learning models, lacking self-learning skills, critical thinking, and the ability to adapt to new
technology - fundamental factors for effectively leveraging the potential of AI.
2.3. Implementation Costs
The initial investment and maintenance costs of AI systems are a major obstacle, especially
for medium and small-sized vocational education institutions. According to the OECD
28
10
5
Tỷ lệ (%)
Có đào tạo bài bản về công nghệ số Từng tiếp cận với ứng dụng AI trong giảng dạy
Có thể chủ động triển khai AI vào dạy học
125
(2023), the average cost to deploy a personalized learning system using AI for a medium-
sized vocational school can be up to $250,000/year - including software, hardware,
personnel training, and technical maintenance costs. Meanwhile, the budget for vocational
education in Vietnam is still limited and mainly focuses on traditional items such as
facilities, enrollment, and teacher training.
The lack of financial support policies from the state or effective public-private partnership
models makes it impossible for most vocational training institutions to proactively access
AI. Even with technology, maintaining the system - including algorithm updates, data
processing, and security - also requires a stable, professional resource that many units
currently are still unable to meet.
2.4. Ethical and Legal arriers
AI in education raises many ethical questions related to privacy, transparency, and
accountability. Research by the AI Now Institute (2022) points out that most AI-based
student assessment systems still lack clear explanation mechanisms about how results are
produced. This can cause suspicion, lack of fairness in assessment, and lead to serious
consequences regarding psychology or career opportunities for learners.
The collection, storage, and processing of student data also pose significant risks without
transparent consent procedures. While countries like Germany, France, and Singapore have
issued specific legal frameworks to regulate the application of AI in education, Vietnam
currently does not have a specific regulatory system, causing difficulties in handling
disputes or complaints related to errors or personal data violations.
2.5. Technology Acceptance
The acceptance of AI by teachers and learners is still not high, reflected in the psychology
of apprehension, worry, and lack of trust in technology. According to a report from the Asia-
Pacific Vocational Education Forum (2023), about 38% of vocational teachers in Vietnam
still believe that AI could reduce their role or even completely replace teaching jobs.
Meanwhile, 44% of students responded that learning with AI lacks human-to-human
interaction, leading to feelings of boredom and lack of learning motivation.
The educational culture in many localities is still heavily traditional, valuing direct
classroom forms and the role of the teacher. This requires communication strategies,
awareness education, and the implementation of successful pilot models to build trust,
encouraging more natural and positive acceptance of AI in the vocational education
community.
2.6. Digital Divide
Finally, the digital divide - manifested through unequal access to technological
infrastructure - is a systemic challenge. According to ITU (2022), only about 64% of
Vietnam's population has internet access, and this rate decreases significantly in high
mountainous regions, islands, and economically disadvantaged areas. Not only lacking
infrastructure, people in these areas also lack technological skills and economic conditions
to access the necessary devices for online learning or AI-integrated training models.