Understanding AI and the human condition

Humans are remarkably capable. We build complex things, solve hard problems, and adapt to circumstances that would have been unimaginable to previous generations.

But capability and confidence are not the same thing, and the gap between them grows wide whenever we encounter something unfamiliar.

Think about assembling flat-pack furniture for the first time. Or changing your own oil. Or fixing a leaking tap when you have never touched a spanner in your life. None of these tasks are beyond the average person’s intellectual capacity.

But if you do not do them regularly, they generate a specific and recognisable feeling: a low-grade, persistent anxiety. You circle the instructions. You second-guess each step. You watch YouTube videos. You wonder whether you are about to make something worse.

That is exactly the psychological state in which most Australian employees are being asked to adopt AI, and we are surprised when they hesitate. The human condition cannot be over-emphasised and should not be underestimated in achieving success.

Exploring AI sentiment in Australia

Australia ranks equal lowest for AI sentiment, trailing 21 other countries, according to the EY Global AI Sentiment Study in 2026. The result highlights a trust gap that could slow confidence in adoption and value creation.

Based on the data, AI use and exposure continue to rise across the country, with 77% of Australians reporting they use AI. However, Australia places 52 out of 100 on AI sentiment – well below the global average of 66 – reinforcing the gap between use and trust.

Key findings include:

• 51% of Australians like the idea of using AI to simplify life and support choices, compared with 64% globally, highlighting more cautious sentiment locally.
• 26% of Australians agree that human oversight is not needed even if AI is accurate. This underscores strong expectations for control and accountability.
• Australians are most comfortable with AI in low-risk, everyday settings such as customer and retail services (61%). However, in higher‑risk areas such as government and financial services, fewer than four in ten Australians report conscious AI use.
• 68% of Australians are worried about losing control over decisions made by AI on their behalf.

This is not a technology problem. It is a human problem. And human problems require human solutions.

The anxiety is not irrational, however.

Asking someone to use a tool they do not understand, within a refined process they cannot fully see, toward outcomes that have not been clearly defined, is a reasonable thing to be anxious about. The answer is not to dismiss the concern, rather to remove its causes.

The first cause is lack of experience.

Anxiety diminishes sharply once people have done something a few times. Experience-based training is therefore critical: not slide decks about AI’s potential, but hands-on practice in context, using tools relevant to the person’s actual role.

The moment someone completes a real task more easily with AI assistance, the relationship with the technology shifts. Familiarity breeds confidence.

The second cause is lack of guardrails.

When people do not know the boundaries, what they are permitted to do, what to avoid, what happens if something goes wrong, they default to caution.

Effective AI governance addresses this directly. It defines how AI is identified, developed, deployed and managed, and it establishes clarity across the five dimensions that determine whether adoption succeeds: strategy linked to business objectives; risk identification across privacy, bias and security; data quality and compliance; infrastructure readiness; and an honest assessment of workforce capability.

These are not sequential steps. They are interdependent, and gaps in any one of them sustain the conditions for anxiety.

Guardrails are not restrictions. They are permissions. They give people the confidence to engage, and they create the evidence trail that leaders and boards increasingly require to trust but verify.

The third cause, and the most corrosive, is the fear of replacement.

No amount of productivity messaging overcomes this if it is left unaddressed. Employees are watching. When organisations lead with efficiency gains and cost reduction, without explicitly framing AI as augmenting human capability rather than substituting for it, anxiety hardens into resistance.

Linking strategy, tech and people

This is where the integration of strategy, technology and people becomes decisive.

Technology is the enabler, not the driver. Strategy defines where AI creates value aligned to corporate objectives. People determine whether any of it works.

Organisations that align these three elements from the outset realise materially better outcomes than those that treat AI as a technology project with change management bolted on at the end.

Bringing people along has an outsized impact on successful implementation, and that means involving employees in design and definition early, communicating transparently about goals and uncertainties, building genuine AI fluency through learning investment, and addressing job displacement fears openly and before they calcify.

Commonwealth Bank recently announced a $90 million workforce investment program to help employees prepare for AI-driven change, developed with unions and designed to give people agency over how their careers evolve. Their AI learning series has reached more than 27,600 employees.

As CEO, Matt Comyn has been explicit that the impact of AI will be uneven and that uncertainty is real.

That is the model all organisations should be following, investment in people before the disruption arrives, and a clear signal that the workforce is part of the solution.

AI strategy must start at the top, owned by the CEO and Executive Team, aligned to business objectives. Identification and definition of the highest material value use cases in the core operations of the business ensures that every person can understand what it means for them.

From that foundation the delivery team can shift to execution seeking to demonstrate results with a consistent focus on enabling and engaging the workforce rather than simply extracting cost. People lean into opportunities when they understand them.

The human condition cannot be over emphasised and should not be underestimated in achieving AI success.

Lindsey Hershman is part of Moxie Top Minds, an editorial-led knowledge network housing the most influential business and technology leaders in Australia. Lindsey is also Managing Director of AI Access, a full-service AI consulting firm specialising in AI Strategy, Governance, Custom AI Development and Adoption.