Organisations Are Teaching AI All Wrong
Most organisations are approaching AI adoption as though it were simply another software rollout. The process is familiar: schedule a training session, teach employees how to write prompts, distribute a policy document, run a webinar, record attendance, and move on. Yet leadership teams are often left wondering why, six months later, AI tools remain underused—or why employees are using them quietly and outside official channels.
The uncomfortable truth is that AI is not primarily a technology problem; it is a human behaviour problem. Many organisations are making the very mistake they hope AI will help them avoid: they are over-engineering the solution while overlooking the human factors that determine whether meaningful adoption occurs.
The Great AI Training Illusion
If you have attended a corporate AI workshop recently, you have probably encountered a familiar format. Participants are shown the best prompts, taught how to write instructions, introduced to document summarisation, and given examples of task automation. None of this is inherently wrong. The issue is that organisations often treat AI as a technical skill that can be mastered in isolation rather than as a capability that must be woven into everyday work.
As a result, employees frequently leave these sessions understanding how AI functions but lacking clarity about how it applies to their specific responsibilities. That distinction matters. People did not become proficient with email because they attended email courses, nor did they master search engines or collaboration platforms through certifications alone. They learned by using those tools repeatedly, in context, while carrying out real work. Capability develops through practice, not exposure.
We Are Accidentally Creating AI Shame
Perhaps one of the most damaging consequences of current AI strategies is the emergence of an unspoken culture of shame around its use. Employees are often given contradictory messages. They are encouraged to use AI to improve productivity, yet warned not to become dependent on it. They are told to embrace the technology, while simultaneously being reminded that the work must remain entirely their own. They are invited to experiment, but also informed that their usage may be scrutinised.
These mixed signals create predictable outcomes. Rather than openly discussing how they use AI, many employees begin hiding it. This is not because they are acting improperly, but because they fear judgment. Some even describe using AI in the same hesitant way people once spoke about using calculators during mathematics exams—as though relying on a tool somehow diminishes the value of their thinking.
The irony is striking. For decades, organisations have encouraged employees to use technology to reduce friction and improve efficiency. Yet now, many are subtly communicating that people should feel guilty for doing exactly that. When fear replaces curiosity, adoption inevitably suffers.
The Obsession With Critical Thinking
Whenever concerns about AI arise, the conversation often turns to critical thinking. The argument is understandable, but it is not always precise. Many AI training programmes imply that critical thinking is at risk because AI can generate answers quickly. However, decades of educational research suggest that critical thinking does not emerge simply through instruction.
Critical thinking develops through experience, reflection, debate, application, and challenge. Paulo Freire argued that education becomes ineffective when learners are reduced to passive recipients of information. Genuine learning occurs through engagement and dialogue. Similarly, Chris Argyris demonstrated that organisations struggle when they focus solely on teaching procedures rather than helping people examine assumptions and question underlying beliefs.
Viewed through this lens, the issue is not that AI diminishes critical thinking. Rather, AI exposes environments where critical thinking was never meaningfully cultivated in the first place. The technology reveals existing weaknesses rather than creating them.
We Are Teaching The Wrong Skills
Most organisational AI programmes focus heavily on technical proficiency. Employees are taught how to prompt effectively, automate tasks, and generate outputs. While these skills have value, they are rapidly becoming commodities. AI interfaces are becoming simpler, tools are growing more intuitive, and the technology itself is increasingly fading into the background.
History shows that when technology becomes easier to use, the source of competitive advantage shifts. Success no longer depends on technical mastery alone. Instead, it depends on judgment. The most important questions are not whether someone can use AI, but when they should use it, when they should avoid it, what risks need consideration, what trade-offs exist, and what outcomes they are trying to achieve.
These are not technical questions. They are judgment questions, and judgment is far more difficult to automate than procedural knowledge.
Donald Schön Saw This Coming
Donald Schön’s work on reflective practice provides a valuable perspective on AI adoption. He argued that professional expertise is not built through memorising rules or following procedures. Instead, expertise develops through reflection while engaged in real work. Professionals become effective because they learn how to navigate uncertainty, adapt to changing circumstances, and make informed decisions in complex situations.
AI adoption follows the same pattern. People do not become effective AI users through theoretical instruction alone. They improve through experimentation, trial and error, iteration, reflection, and conversations with colleagues about what works and what does not. Learning emerges from doing, not merely from being told.
AI Should Be Like Electricity
The most transformative technologies eventually become invisible. Nobody announces in a meeting that they used electricity to write a report. Few people celebrate their use of word-processing software or seek recognition for sending an email. These technologies became powerful precisely because they disappeared into the background of everyday work.
The same is likely to happen with AI. The organisations that thrive will not necessarily be those that talk about AI the most. Instead, they will be the ones that quietly integrate it into workflows, decision-making processes, learning systems, customer service, analysis, and creative work. In those organisations, the technology fades into the background while the work remains at the centre of attention.
What Organisations Should Do Instead
Rather than investing exclusively in more training programmes, organisations should create more opportunities for experimentation. Instead of relying on policies that generate anxiety, they should focus on building psychological safety. The key question is not whether employees know how to use AI, but whether they feel safe enough to use it openly and responsibly.
Success should not be measured solely through attendance figures or completed training modules. It should be measured through meaningful adoption and practical application. Likewise, instead of appointing isolated AI champions, organisations should foster communities of practice where employees can learn from one another, share experiences, and collectively solve problems.
Research on workplace learning consistently demonstrates that professional capability develops socially through observation, discussion, collaboration, and shared problem-solving. Formal training has a role to play, but it is rarely sufficient on its own. The same principle applies to AI.
The Real Competitive Advantage
Many organisations assume their competitive advantage will come from having access to the best AI tools. That assumption is increasingly difficult to defend as powerful technologies become widely available. The real advantage is likely to come from something far more human.
Organisations that cultivate curiosity, encourage experimentation, and create environments where learning is openly shared will be better positioned to succeed. Leaders who model exploration, teams that exchange insights, and employees who exercise sound judgment will create value that competitors cannot easily replicate. Most importantly, organisations that make people feel safe enough to experiment without fear of embarrassment or punishment will unlock far greater potential from AI than those focused solely on technical implementation.
Ultimately, the future of AI adoption is not about becoming more technical; it is about becoming more human. That may be the greatest irony of all. The technological revolution dominating today’s conversations may ultimately teach us less about artificial intelligence and far more about organisational psychology, culture, and human behaviour.
