As AI becomes more capable, much of the discussion has focused on which jobs it might replace.
But a job is rarely one task. It brings together responsibilities, decisions, relationships and activities, and AI will affect each of these differently. Some tasks may become automated, while others may require more judgement, context and human input.
That means the impact of AI can be felt within a role long before the role itself disappears. The title may stay the same while the work, skills and capabilities required to perform it begin to change.
For organisations, the challenge is recognising these shifts early enough to understand what they mean for roles, skills and workforce requirements.
AI Is Changing the Composition of Work
AI can already handle a growing range of activities, from processing information and analysing data to generating content and supporting routine decisions. As these capabilities become part of everyday work, the balance of responsibilities within many roles will begin to shift.
Automating a task does not necessarily remove the role responsible for it. In many cases, it changes where that person spends their time and where their judgement is most valuable.
Take a finance professional. AI may handle more reporting, data processing and forecasting, while the role places greater emphasis on interpreting information, challenging assumptions, advising the business and assessing different scenarios.
The job still exists. What has changed is the work within it.
Job Titles Do Not Tell the Whole Story
One of the less visible effects of AI adoption is that job titles can stay the same while the capabilities required to perform those roles change significantly.
A marketing role may involve less manual content production and more strategic direction. An analyst may spend less time preparing data and more time interpreting it. A customer service professional may handle fewer routine enquiries while taking on more complex conversations that require judgement and relationship management.
The title may look familiar, but the work behind it can be very different.
That makes job titles an unreliable measure of how a workforce is changing. Organisations need to look more closely at the work itself: which responsibilities are being automated, which are becoming more important, and which new capabilities are emerging.
These shifts can reveal changes in workforce capability that headcount alone cannot show.
Redesigning Roles Around the Work
Once the work inside a role changes, the organisation has to decide how the role should evolve with it.
That can affect more than the responsibilities attached to one position. Changes to a role can alter team structures, reporting lines, decision-making and the capabilities needed elsewhere in the organisation.
The challenge is making those changes deliberately. Without a clear understanding of how work is evolving, organisations can introduce new technology while leaving roles, responsibilities and team structures largely unchanged.
Job redesign therefore needs to be considered as part of workforce planning, rather than treated as a separate technology exercise.
The Capability Gap
The bigger workforce issue may not be having enough people, but having the right capabilities for the work that needs to be done.
Some people will be able to move into redesigned roles through learning and development. Others may need new tools, different responsibilities or opportunities to build complementary skills. In some cases, the capabilities required may not exist internally at all.
That creates different options for workforce strategy. Organisations may need to develop existing talent, hire new expertise, bring in specialist capability for a defined period, or use more flexible talent models while longer-term requirements become clearer.
The right approach will depend on the role, the organisation and how quickly the work is changing. What matters is recognising that AI adoption has direct implications for workforce capability. Decisions about technology, roles and talent therefore need to be considered together.
Workforce Planning Has to Reflect the Work
Understanding how work is changing gives organisations a better basis for workforce planning.
Instead of making headcount decisions based on what AI could automate, organisations can assess how roles are likely to evolve, where existing capability can be redeployed or developed, and where new expertise will be needed.
That creates a clearer connection between technology investment and workforce decisions. It also reduces the risk of removing capacity simply because part of a role can be automated, without considering the work and capability the organisation will still need afterwards.
The goal is not to predict exactly which jobs will remain unchanged. It is to understand what the organisation will need as the work continues to evolve.
The Workforce Will Evolve Alongside AI
AI adoption will have implications beyond individual roles, influencing how organisations structure teams, allocate responsibility and plan for the capabilities they need. Workforce planning therefore needs to account for changing capability requirements rather than treating technology investment and talent decisions as separate priorities.
The organisations that respond well will connect technology decisions with the workforce required to make them effective. They will understand where technology can improve how work gets done, where human judgement remains valuable, and where their workforce needs to develop, adapt or be supplemented with new expertise.
AI does not determine the shape of the future workforce on its own. Organisations still have to make deliberate decisions about the work they need, the capabilities behind it and how those capabilities will be built.
The real test is whether the organisation can build the capability those changing jobs will require.
Job redesign needs to be considered as part of workforce planning, rather than treated as a separate technology exercise.