Currently, 43% of Thai businesses use AI on an ongoing basis, but only 4% have achieved full adoption, with the number one obstacle being a "lack of employee skills". Upskilling the entire organisation in AI should therefore not start with sending staff to scattered tool-based classes, but by dividing AI skills into three tiers based on job roles, then planning targeted development for each group with clear measurement. This article summarises a 3-tier AI skills framework and an actionable roadmap for tangible results.
Hello. Lately, Lyn has been hearing the same question from HR teams across many organisations: "Management wants all employees across the organisation to be proficient in AI, but where do we start? Should we send everyone to learn ChatGPT, or just buy online courses for self-study?"
Lyn completely understands how pressing this is, because the urgency is very real. According to AWS's "Unlocking Thailand's AI Potential 2026" report, Thai businesses using AI continuously increased to 43% from 32% the previous year; among active adopters, 84% reported improved productivity, and 71% saw an average revenue increase of 19%. This means early-adopting organisations are pulling ahead faster and faster.
But the answer to "where to start" is not "send everyone to study ChatGPT". Real, impactful organisation-wide AI upskilling starts with categorising AI skills into three tiers based on job roles: Assistant tier (AI Literacy and Copilot), Specialised Application tier (AI Application), and Strategic tier (AI Strategy), followed by designing targeted development tied to real tasks and clear measurement. Today, Lyn will walk you through each step.
Why does letting employees learn AI on their own fail to work?
Because tool access does not equate to effective workplace application. Employees might watch YouTube tutorials or experiment with ChatGPT independently, but without designing learning that connects to their daily tasks and sets shared standards, everyone ends up doing their own thing—some become proficient, while others never touch it at all.
The figures clearly reflect this. Despite high AI awareness in Thailand, a report by Telenor Asia indicates that only around 20% of Thai workers actively use AI tools in their day-to-day work. The gap is not about "knowing AI", but rather "knowing how to apply it effectively to one's job".
This is an organisation-wide challenge, not merely an individual effort. The Microsoft Work Trend Index 2026 survey highlights that culture, leadership, and people development are twice as influential on AI success as individual efforts alone. Leaving employees to figure it out on their own therefore fails to address the root issue.
Where should you start when developing AI skills across the organisation?
It starts by acknowledging that not everyone needs the same training. Frontline staff, team leaders, and executives each require distinct AI skill sets. Effective planning begins by segmenting employees by role and defining what each group needs AI to accomplish in their actual work, rather than jumping straight into selecting courses.
Data strongly supports this. The Microsoft Work Trend Index 2026 proposes a 3-tier roadmap: employees develop critical thinking and verify AI outputs, leaders embed AI into core job responsibilities, and organisations track internal AI adoption to redesign workflows. This aligns seamlessly with how Learning Hub Thailand structures its AI for Business category courses, segmented precisely into three levels of depth.
The urgency is undeniable. An Ecosystm report estimates that Thailand faces a shortage of around 80,000 AI-proficient personnel across all business functions, not just technical fields. Upskilling your existing workforce is therefore far more cost-effective and faster than hiring from scratch.
Understand the 3 Tiers of AI Skills Before Planning
The compass Lyn recommends HR teams use is viewing AI skills across three tiers and mapping each employee group accordingly:
Level 1: AI Literacy and Copilot (For All Employees) Using AI as an assistant for routine tasks, such as summarising documents, drafting emails, and conducting preliminary data analysis. The goal is to help everyone work faster without changing existing work systems. This group is ideal for courses such as AI Copilot for Business Productivity and Impact and Advanced ChatGPT: From Fundamentals to Pro Level
Tier 2: AI Application (For specialised teams) Diving deep into applying AI to domain-specific work, such as data analysis and data storytelling. This tier suits courses like Data Storytelling with AI Copilot, which helps transform raw data into sound business decisions.
Tier 3: AI Strategy and Team Leadership (For leaders and executives) Understanding the big picture of how AI reshapes business processes and strategies, alongside managers leveraging AI for team administration. This tier is designed for AI for People Managers and AI Strategy and Transformation for Executives.
A critical element that must never be overlooked is cultivating "critical discernment" alongside tool usage across all levels. According to the Microsoft Work Trend Index 2026, only 45% of Thai professionals prioritise critical thinking when working with AI—the lowest in ASEAN—while only 53% focus on AI output quality verification. Teaching staff merely how to click buttons without verifying outputs yields speed at the expense of quality and risk.
Assess Where Your Organisation Should Start
Before examining the step-by-step roadmap, Lyn invites you to assess where most of your staff currently stand. In some organisations where employees have never touched AI, starting comprehensively at Tier 1 is essential. In others where staff are already adept but leadership lacks strategic vision, strengthening Tier 3 is the priority.
To explore what each tier covers and which groups it suits best, take a look at the overview of AI for Business courses across all 3 tiers, and consult our team to tailor the content to your organisation's context.
Step-by-Step AI Upskilling Roadmap (For HR)
You can structure the plan into five manageable steps without trying to do everything all at once, achieving tangible results group by group:
1. Assess AI readiness first Because organisational readiness varies significantly: the AWS 2026 report found that 47% of startups are AI-ready, compared to 24% of large enterprises and 17% of SMEs. Knowing your true baseline prevents setting expectations too high or too low.
2. Select a pilot group There is no need to roll it out organisation-wide simultaneously. Choose one or two teams with high volumes of repetitive work and quick-win potential to create success stories that inspire others.
3. Select skills based on actual workflows, not trends Identify specifically where AI can assist this group's daily work and choose courses that address those exact needs, rather than training just because a topic is trending.
4. Design hands-on practice with real-world tasks In line with AWS's recommendation that training should emphasise applied critical judgement rather than basic digital literacy, encourage learners to bring their own real work to class so they can apply it from day one.
5. Measure and scale Define metrics from the start, such as hours saved or volume of AI-assisted outputs, then use the pilot group's results to scale across subsequent teams.
The Difference Between Ad-hoc Training and a Systematic Organisation-Wide Approach
Here is a clear comparison:
Ad-hoc Training
Systematic Organisation-Wide Approach
Sending everyone to the same course
Segmenting skills into 3 tiers based on job roles
Choosing courses based on hype and trends
Selecting skills tailored to each group's real work
Everyone working in silos after the training
Establishing shared standards and practices across the organisation
Teaching only tool mechanics and button-clicking
Teaching tool mastery alongside output verification
Measured by the number of people trained
Measured by transformed workflows and time saved
This distinction is vital. Although Thai businesses are increasingly adopting AI, AWS 2026 found that among those familiar with AI, only 4% have fully integrated it into their operations. The gap between 'trying it out' and 'transforming the organisation' lies in system planning, not having countless courses.
Workplace Stories: From a one-off AI workshop to practical team adoption
Here is a common scenario I often observe across various organisations, rather than any single individual.
Many HR teams share a similar story: they organised a company-wide, one-day ChatGPT training session. Everyone had great fun on the day, sharing photos in their LINE groups. Yet a month later, when asking who was still using it, only a handful remained. People had simply reverted to their old routines because no one helped bridge the gap between AI and their daily tasks.
When they changed their approach—without increasing the training budget, but shifting their planning—they started with a marketing team handling repetitive content tasks as a pilot group. They brought actual work into the classroom and set clear targets to measure time saved. As a result, this team adopted AI consistently and became a role model for other departments. The key takeaway: what makes AI stick across an organisation is not the number of people trained, but planning so each group applies it to real work and achieves tangible results.
Reflection Questions: Food for Thought
- Do you segment employees by role before choosing courses, or send everyone to the same training?
- Are employees who completed training still applying AI to their daily work, or have they reverted to old habits?
- Do you teach employees to review and verify AI outputs, or just how to use the tool?
- Do you have a pilot group with clear results to serve as a model for other teams?
- Do you measure AI training success by workflow improvements, or simply by attendance numbers?
Next Action: Start today
If you want to start planning an AI upskilling initiative, I recommend starting with small, actionable steps this week. Choose just one team with the highest volume of repetitive tasks, and identify 3 tasks where AI could help. Then set a modest goal for this team to apply AI to those 3 tasks within the month, and measure the time saved.
With just this, you will have a tangible pilot group, rather than running company-wide training and hoping people continue to use it. I believe organisations where entire teams leverage AI effectively do not start with the largest number of courses, but by planning skills tailored to roles and enabling everyone to apply them to real work. 💛
Ready to plan an organisation-wide AI upskilling programme?
Learning Hub Thailand offers comprehensive AI for Business courses across all 3 levels, from AI Literacy and Copilot for all staff, to AI Strategy for executives. Designed with a Learning by Doing approach, content can be tailored to match your organisation's real-world workflows.
- 👉 Explore the complete overview of AI for Business courses and consult our team to design an AI upskilling plan tailored to your organisation.
- 👉 If you want employees to quickly start using AI for their daily work, start with AI Copilot for Business Productivity and Impact or Advanced ChatGPT from Basics to Pro.
Frequently Asked Questions
Where should we start when upskilling AI skills for employees across the organisation?
Start by categorising employees by role and assessing AI Readiness, rather than simply choosing a course. Next, select a pilot team with high volumes of repetitive tasks and quick wins. Have them bring real work to test with AI during the training, and measure the results before expanding to subsequent groups. This approach delivers tangible results much faster than rolling out training across the entire organisation at once.
Can employees without an IT background learn AI?
Yes. Most corporate AI courses are designed for general business professionals without requiring any coding. They focus on using ready-made AI tools controlled via natural human language, such as ChatGPT, Gemini, and Copilot. Learners will practise effective prompting and apply it directly to their real work tasks during the workshop.
How many levels should employees' AI skills be divided into?
Divide into 3 levels based on job roles: Level 1 AI Literacy and Copilot for all employees to use AI as a daily work assistant; Level 2 AI Application for specialised teams, such as data analysis and storytelling; and Level 3 AI Strategy for managers and executives who need to understand the strategic overview. Crucially, all levels must incorporate training on validating AI outputs.
How does In-house Training for AI differ from having employees learn online independently?
Self-paced online courses or YouTube videos provide generic knowledge for everyone, but they are not tailored to an organisation's specific workflows. In-house Training brings real tasks, real tools, and real challenges into practical workshops, enabling employees to apply AI to their own work from day one whilst establishing shared AI standards across the organisation.
How can we measure the impact of employee AI training to see real results?
Measure changes in actual work and tangible outcomes, not just attendance numbers. Examples include hours saved per week, the number of tasks supported by AI, or improved work quality. Define these metrics before the training begins and evaluate the pilot group's results to inform decisions on expanding to other teams.
References:
- AWS – Unlocking Thailand's AI Potential 2026 (via SME Asia): 43% of Thai businesses use AI, but only 4% have achieved full deployment.
- Microsoft – Work Trend Index 2026 (via The Story Thailand): 32% of Thai workers are advanced AI users + a 3-tier roadmap.
- Ecosystm – AI in Thailand: Enterprise Adoption & Readiness (facing a shortage of around 80,000 AI professionals)
- Nation Thailand – Lack of AI skills threatens Thailand's global standing (around 20% of Thai workers use AI at work)
