Everyone Is Training on AI. Almost Nobody Is Building AI Capability

Why AI awareness is not enough—and what institutions must do next

Dr. Sheila Jagannathan
Vice President, Strategic Relationships and Outreach, Saylor University; Senior Advisor on 21st Century Skills, Digital Learning and Capacity Development to the African Development Bank and World Bank; Former Global Head, Open Learning Campus, World Bank

Every week, another university launches an AI certificate. Another government announces an AI skilling initiative. Another company rolls out mandatory AI training for employees. In boardrooms, ministries, and universities around the world, AI training has become one of the hottest investments in workforce development.

The numbers are impressive.

Thousands of AI courses are now available online. Millions of learners have enrolled in AI-related programs. Governments across Africa, Asia, Europe, and Latin America have announced national AI strategies that place digital skills and AI literacy at the center of workforce development.

Yet a fundamental question remains: Will all this AI training actually lead to AI capability?

Increasingly, I believe the answer is NO.

After more than three decades working at the intersection of learning, technology, workforce development, and institutional capacity building—including leadership roles at the World Bank, the African Development Bank, and now Saylor University—I have observed a recurring pattern.

Organisations often confuse training people about technology by helping people use technology effectively.

They are not the same thing.

I have seen this before with e-learning, MOOCs, knowledge management platforms, and digital transformation initiatives. Technology adoption rarely fails because of technology. It usually fails because organisations underestimate the behaviour, institutional, and cultural changes required to translate learning into performance.

Today, we risk repeating this mistake with artificial intelligence.

The Great AI Training Rush

The current enthusiasm for AI training is understandable.

Artificial intelligence is advancing at remarkable speed. Tools such as ChatGPT, Claude, Gemini, Copilot, and others are reshaping how people search for information, write reports, analyse data, create content, and solve problems.

Universities are rushing to create AI courses.

Governments are investing in national AI literacy programs.

Businesses are training employees to use generative AI tools.

These efforts are valuable and necessary.

But they are only the first step.

Much of today’s AI training focuses on awareness:

• What is AI?
• What is generative AI?
• How do prompts work?
• What are the risks?
• What are the opportunities?

These are important questions.

However, knowing the answers does not necessarily change how people work.

A two-hour webinar on AI will not transform a ministry.

A short online course will not modernise a university.

An AI certificate alone will not make an organisation or an individual AI-ready.

Awareness Is Not Capability

Capability is different.

Capability means that people can consistently use AI to improve performance, productivity, decision-making, innovation, and service delivery.

For example, a ministry may train hundreds of staff on generative AI. Yet unless policy analysts use AI to improve research, managers redesign workflows, and teams integrate AI into their daily operations, organisational performance may remain largely unchanged.

Capability exists when:

• Teachers redesign learning experiences using AI.
• Civil servants improve policy analysis using AI tools.
• Project managers use AI to accelerate reporting and monitoring.
• Entrepreneurs use AI to expand businesses and reach new markets.
• Researchers use AI to analyse large datasets more effectively.

In other words, capability emerges when learning translates into action.

Unfortunately, many organisations stop at awareness.

They provide courses but do not create the conditions necessary for adoption.

The result is predictable.

Employees complete training.

Certificates are issued.

Dashboards show impressive participation rates.

Yet little changes in daily practice.

Figure 1. From AI Awareness to AI Capability

The Missing Piece: Organisational Adoption

The real challenge is not learning AI.

The real challenge is integrating AI into how work gets done.

This requires much more than training.

It requires changes in behaviours, workflows, leadership, governance, and culture.

Behaviour and Workflows

Are the incentives right for changing attitudes toward AI adoption?

Does AI simplify routine tasks for significant cost savings?

Which processes should be redesigned?

Where can automation reduce administrative burdens?

Leadership

Do leaders actively encourage experimentation?

Are managers modeling AI use themselves?

Do employees feel safe testing new approaches?

Governance

What policies guide responsible AI use?

How are privacy, ethics, and data security managed?

Who is accountable?

Culture

Are people rewarded for innovation?

Do organisations encourage learning from failure?

Is continuous learning embedded into the workplace?

Without these elements, AI training remains largely theoretical.

The Learning Ecosystem Advantage

One of the most important lessons I have learned from building large-scale learning programs is that courses alone are never enough.

Learning happens within ecosystems.

Over the years, institutions have invested heavily in Learning Management Systems (LMSs), course libraries, and digital content repositories. Many believed that if they acquired the right technology platform, learning would naturally follow.

Experience has shown otherwise.

Technology matters. Content matters. But sustainable learning outcomes typically emerge when institutions combine these investments with communities, incentives, leadership support, mentoring, practical application, and ongoing performance support.

Successful learning ecosystems include:

• Applied projects
• Communities of practice
• Peer learning
• Coaching and mentoring
• Knowledge-sharing platforms
• Performance support tools
• Digital credentials
• Analytics and feedback loops

The future belongs to organisations that connect these components into coherent learning journeys.

This is especially true for AI.

AI capability cannot be developed through isolated courses alone.

It requires ongoing practice, experimentation, collaboration, and support.

A Lesson from Open Learning

One encouraging development is the expansion of free and accessible learning opportunities that enable learners around the world to acquire new skills regardless of geography, income, or institutional affiliation.

At Saylor University, where I currently serve as Vice President for Strategic Relationships and Outreach, we are seeing growing demand for free and flexible learning pathways in AI, data science, business, entrepreneurship, and workforce skills. Through our open learning model, learners from more than 190 countries can access over 150 university-level online courses at no cost, helping to democratize access to high-quality education and lifelong learning.

What is particularly striking is that learners are increasingly asking different questions than they were even two years ago.

Rather than asking, “What is AI?”, they are asking:

“How can AI help me become a better manager?”

“How can AI improve my productivity?”

“How can AI help me start or grow a business?”

“How can AI support teaching, healthcare, research, or public service?”

In short, they are seeking AI capability rather than AI awareness.

This shift is also reflected in the evolution of learning offerings. In addition to its free course catalogue, Saylor University now offers graduate degree programs, including an MBA, Master of Management, Master of Marketing, and Master’s in Entrepreneurship and Innovation, while also developing a master’s degree in Artificial Intelligence to respond to growing workforce demand.

What we are observing reflects a much broader trend across higher education and workforce development.

People are not looking for AI knowledge in isolation. They are looking for practical ways to apply AI to improve their work, advance their careers, solve problems, and create value.

They are looking for AI capability in context.

Explore Saylor University’s free course catalogue: https://learn.saylor.org

Graduate degree programs: https://degrees.saylor.org

This distinction between awareness and capability may ultimately become one of the defining challenges—and opportunities—of the AI era.

Five Principles for Building AI Capability

Based on lessons learned across governments, universities, development organisations, and large-scale learning platforms, I believe five principles are essential.

1. Move Beyond Awareness

AI literacy is an important starting point, but it is only the foundation. Too many organisations stop after introducing staff to AI concepts and tools.

The objective should be to help people apply AI to real-world challenges. For example, rather than simply teaching public servants how generative AI works, organisations should help them use AI to draft policy briefs, summarise consultations, analyse stakeholder feedback, or improve citizen services. Similarly, educators should move beyond explaining AI and focus on redesigning teaching, assessment, and learner support.

Awareness creates understanding. Application creates capability.

2. Redesign Work, Not Just Training

Many organisations focus on training programs while leaving existing workflows unchanged.

The real opportunity lies in redesigning how work is performed. For example, project teams can use AI to accelerate proposal development, generate first drafts of reports, summarise meeting discussions, or analyse large volumes of data. Universities can use AI to support curriculum design, personalised learning, and student advising. Governments can use AI to streamline administrative processes and improve service delivery.

The goal is not more learning activity. The goal is to improve performance.

3. Build Communities of Practice

People rarely change behaviour because of a course alone. They change when they see peers successfully applying new approaches.

Communities of practice allow practitioners to share use cases, lessons learned, prompts, tools, and emerging practices. In many organisations, the most valuable learning occurs when colleagues demonstrate how they are using AI to solve real problems.

The most effective AI learning initiatives combine formal learning with peer learning, collaboration, mentoring, and knowledge sharing.

4. Create Continuous Learning Pathways

Unlike many technical skills, AI is evolving almost weekly. A course completed six months ago may already be outdated.

Organisations therefore need to move from one-time training events to continuous learning pathways that combine short courses, practical assignments, webinars, communities, performance support resources, and ongoing experimentation.

Learners need opportunities to continually update their skills as tools, capabilities, and use cases evolve. In the AI era, learning is no longer an event. It is a continuous process.

5. Measure Impact, Not Participation

Many learning initiatives celebrate the number of people trained, courses completed, or certificates awarded.

While these metrics are useful, they do not tell us whether learning has made a difference.

The more important questions are:

• Are employees using AI in their daily work?
• Has productivity improved?
• Have processes become faster or more efficient?
• Are better decisions being made?
• Has innovation increased?

Success should ultimately be measured not by learning activity, but by organisational outcomes.

Organisations that focus on these five principles will be far more likely to move beyond AI awareness and develop the capabilities needed to thrive in an increasingly AI-enabled world.


Written for OEB 2026.