Authored by Lindsay Ratcliffe, chief innovation & transformation officer at CreateFuture

A series of misnomers are increasingly shaping conversations on the impact of AI on work. That AI adoption will diminish people’s roles and capabilities, that automated productivity automatically creates value, and that successful transformation is driven by the technology rather than the choices organisations make around it.

These are natural assumptions, but there is a sense that it is creating unnecessary anxiety among employees and leading businesses to approach AI in the wrong way.

The reality I encounter with clients is far more nuanced. The scale of shadow IT in many organisations suggests this is rarely a technology being forced upon an unwilling workforce. More often, employees are already experimenting with it, informally and ahead of the policies intended to govern its use. 

That leaves leaders and employees in an awkward situation. People are using AI, but often without much clarity about what it means for their work or where it leaves them. Leaders therefore need to move beyond broad promises about productivity and have a more honest conversation about how roles, skills and organisations will change.

‘AI will inevitably erode our skills’ 

A recent MIT study warned that over-reliance on AI tools can create ‘cognitive debt’. I’ve heard these concerns first hand from senior engineers who feel they’ve gone from writing code to reviewing outputs, and have lost some of the craft they valued in their work.

But there is nothing inevitable about AI making people lazy. The risk arises when it becomes the default response to every task, including the work through which people develop their judgement and expertise. 

This is why blanket policies tend to be unhelpful. The more useful distinction is between work where AI genuinely adds value and the work people need to do for themselves. No business wants to outsource so many skills – from email diplomacy to design and coding – that its people can no longer work effectively without AI. 

AI use therefore has to be considered rather than automatic. People need a clear sense of which skills still matter and where human judgement takes precedence. 

‘AI will limit career opportunities’ 

This concern is especially acute among younger workers, who are wondering not only what their careers might look like, but whether they’ll get a foot on the ladder at all. 

Simply saying that businesses will always need fresh thinking doesn’t really address that concern. If AI takes on much of the routine work traditionally done by junior employees, how will those employees gain the experience they need to make more complex judgements later?

It’s a question I don’t think many organisations have fully considered yet. 

Take a graduate business analyst. Traditionally, they might sit in stakeholder interviews, map processes and pull together the first draft of requirements. AI can do much of that work far faster. But by doing those tasks themselves, the graduate wasn’t just simply producing output but learning. Learning how to ask better questions, learning how to spot contradictions, and understanding why different stakeholders see the same problem differently and recognising when something doesn’t quite add up.

That doesn’t mean we should just preserve routine work simply because that is how people used to learn. But organisations can’t assume that by asking junior employees to check AI-produced work that they will develop the same skills and judgement.

If AI removes part of the old “apprenticeship model”, organisations will have to consciously design a new, AI-augmented one. That means considering AI transformation and talent development as one – what should be automated, what should people learn by doing, and how do you develop judgement in an AI-driven environment.

The concern isn’t simply whether AI will reduce the number of junior roles today, but whether organisations will still have the experienced people they need tomorrow. Recruitment, learning and development, and transformation teams therefore need a shared understanding of where the next generation fits.

‘AI is just a faster horse’ 

AI can resolve known technical issues and modernise ageing IT architecture far faster than human engineering teams. But the bigger opportunities often lie elsewhere: revealing risks buried in messy data, or identifying patterns people could not reasonably detect themselves. 

When I assess an AI programme, the central question is whether it can produce an outcome that is faster, leaner or better, and therefore create more value than would be possible without AI. These are distinct outcomes, and conflating them is where many transformation efforts go off-kilter.

That distinction matters to employees. If every AI initiative is presented as a way to do the same work with fewer people, they will naturally assume efficiency is the only outcome the business values. 

But that is only one application of AI. It can also help people make better decisions or tackle problems that were previously too complex, leaving more room for work that depends on human experience and imagination. The technology required to achieve that can be difficult. But the most important strategic decision is what work should be automated, augmented or orchestrated by AI, and what should remain human.

Those choices have to be specific. What role is AI actually being asked to play? And crucially, what happens to the person whose work is being affected?

Employees don’t need platitudes or empty reassurance that AI is nothing to worry about. They need an honest account of how it is likely to change their work and how they will be able to develop as a result. Getting that balance right is how businesses will create better outcomes while bringing their people with them.