AI readiness goes beyond training numbers: AVPN’s Young Park on building an AI-ready workforce

As organisations across Asia-Pacific accelerate AI adoption, building an AI-ready workforce will require more than expanding access to training and certification courses. Young Park, Director of the AI Opportunity Fund Project at AVPN, argues that readiness should be measured by what people are able to do with AI after training, from improving productivity and income to saving time and applying the technology responsibly in their work.

In this exclusive conversation with us, Park discusses why AI adoption looks different across workers, industries and markets, and why a one-size-fits-all approach risks leaving people with structural barriers further behind. She also looks at the specific challenges facing MSMEs, the role of trainers in a rapidly changing AI environment, and why governments, employers and the wider learning ecosystem need to move from basic AI literacy towards more contextualised, practical and inclusive pathways to AI adoption.

Many organisations, including governments, support workers, MSMEs and different segments of society to build AI skills. But we believe AI readiness should be measured by more than the number of people who complete training. What matters is whether that training leads to changes in confidence, knowledge and, ultimately, behaviour.

The first step is to capture these changes through surveys and self-reported feedback, looking at how people perceive their AI readiness in their daily lives and at work. But that is only the starting point. We also want to understand what happens three or six months after the training, when we can look at more tangible outcomes.

For example, we track individual stories and evidence of changes in income and productivity to understand how people are applying what they have learned. One beneficiary in India participated in our training and reported greater confidence and knowledge afterwards. A few months later, she told us she had applied those skills to her own business, including rebranding and revisiting her pricing strategy. She increased the prices of some products while continuing to see strong customer interest, and her income more than doubled.

In another case, a teacher in South Korea told us she was spending a significant amount of time on lesson planning. After applying the skills she gained through AI training, she was saving around an hour to an hour and a half each day. That gave her more time to spend interacting with her students.

These examples help us understand what AI readiness actually looks like across different contexts, jobs and individuals. Rather than focusing only on how many people have completed training, we should look at the real-world changes that follow, whether that is greater confidence, improved productivity, higher income or time saved.

To understand AI training needs and adoption, we first need to look at what it really means to apply AI at work. Today, if someone uses Gemini, ChatGPT or other AI-powered tools to draft emails, run calculations or complete other everyday tasks, they might simply say, “Yes, I use AI at work.” But that does not necessarily mean they understand how to apply AI effectively, strategically or responsibly in their role.

Meaningful AI adoption in the workplace needs to go beyond that. It involves understanding how to use AI in a more structured, responsible and contextualised way.

For example, when using a simple AI tool to draft an email, employees need to understand what information they can safely share and what sensitive company or work-related information should not be shared. Having access to an AI tool does not necessarily mean people understand how to use it responsibly.

Before addressing barriers to AI adoption, we need to provide well-designed, contextualised training that helps people understand how to use AI effectively and responsibly in their work.

But there is also an awareness gap. Our research found that only 15% of workers surveyed had participated in formal training on applying AI at work, while more than half did not know that free AI learning opportunities were available.

This needs to be addressed across Asia-Pacific, with training tailored to different groups, particularly MSMEs. These businesses often have limited time and resources, so they cannot afford to take employees away from their work for days of training.

We therefore need specialised MSME learning programmes that bring training to them, in formats that fit their time, resources and business realities.

To begin with, training is an important mechanism for MSMEs to give people the knowledge and capabilities they need to grow or start a business and compete on a more level playing field.

But training alone cannot provide the cash investment an MSME may need to purchase inventory, nor can it provide access to digital infrastructure, reliable internet or devices. Those challenges require broader investment and collaborative efforts involving governments, businesses, philanthropies and other stakeholders.

Training can give MSMEs the capacity and confidence to use the resources available, but many business owners need practical guidance on where to start and how to apply AI to their businesses.

In the MSME training sessions I have attended, a recurring question is – “I understand what this tool can do, but what should I actually do with it?” They also have concerns about security, data protection and how to introduce AI safely.

We therefore need more customised training, one-on-one mentoring and practical playbooks that guide businesses through adoption step by step, from understanding data protection requirements to identifying the right tools and asking the right questions before implementing them.

This kind of hands-on support can help MSMEs move from knowing about AI to actually using it, while broader stakeholders address the financial, infrastructure and ecosystem barriers.

When scaling up AI capabilities, the fundamental challenge is the same for HR professionals, L&D teams and AI trainers across the workforce development ecosystem. The business environment is changing very quickly, and these changes can be highly specific to an industry, market or even an individual company.

Trainers therefore need to be open, flexible and adaptable enough to understand these changes almost in real time. I know that is challenging, particularly because AI is evolving faster than many other areas today, but they need to continuously reflect those changes in how they teach and guide their clients or learners. The content itself also needs to evolve accordingly.

At the same time, I don’t think trainers should focus only on new technology. If they do, they can become overwhelmed very quickly. There will always be new models, tools and capabilities being introduced.

Instead, we should start with the fundamental problems that companies and learners are trying to solve. Those problems often change much more slowly. A company may want to engage with potential customers more effectively. A sales team may want to improve how it converts leads into tangible opportunities. An HR team may want to identify the right talent more effectively during hiring.

Those challenges continue to exist, even as the technology changes. So the role of trainers is to start with those problems and identify how new technologies can provide better, more customised solutions.

For trainers, particularly those working in corporate learning and human resources development, that ability to stay adaptable while remaining focused on real business and workforce needs will be increasingly important.

Firstly, a one-size-fits-all approach does not work, particularly when it comes to workers and communities at risk of being left behind.

Different groups face different barriers. This could include people with disabilities, neurodivergent workers, people from different linguistic or cultural backgrounds, or workers living in rural and underserved communities. The first step is therefore to understand what kind of learning environment allows each group to learn more effectively and apply those skills.

That is also where the AI Opportunity Fund started when we launched it in 2024. We wanted to make access to AI learning more available to workers facing structural barriers. Our grantee organisations, for example, have supported single mothers in Japan, farmers in rural areas, people with disabilities in Indonesia, and young people experiencing social isolation in Japan.

We provided a common set of principles and learning objectives, but allowed each organisation to customise the content and training approach around the needs of its particular beneficiary group. I think the same principle applies to employers and HR leaders.

Organisations need to start by understanding who their employees are and what specific barriers different groups may face. They can also learn from the communities and organisations they already work with to understand what challenges people are experiencing and what support they need next.

One example from Singapore really stood out to me. One of our grantees had trained many marginalised communities and, through that experience, started thinking about what should come next. They have now designed an apprenticeship programme to provide work-based learning and project experience after participants complete their training, working with employers and MSMEs.

AI readiness should not end with training, and we also need to think about how people can translate those skills into real work experience and opportunities. Supporting those kinds of pathways will be important if we want AI adoption to be more inclusive.