AI should redesign work—not simply speed up tasks

The most useful starting point for AI is not, “Where can we put AI?” It is, “Where is work repetitive, slow or unnecessarily difficult—and what would better work look like?” Technology is an enabler. The value comes from changing the way work happens and helping people adopt that change.

There are sensible low-hanging-fruit opportunities: recurring queries, first drafts, routine summaries, information retrieval and other repeatable activity. Automating those tasks can release time quickly. But if the organisation stops there, it may make an inefficient process faster without improving the outcome.

Move from task automation to work redesign

Look at the whole flow of work. What triggers it? Who contributes? Where are decisions delayed? Which hand-offs create friction? What judgement must remain human? AI may remove a step, change the information available to a manager or shift responsibility between roles. Those changes affect capability, accountability and employee experience—not only productivity.

This is where HR and organisation development have a central role. Leaders need to understand how roles will evolve, which skills become more important and how performance should be assessed when people are working with intelligent tools. Employees need clarity on what good use looks like, where the boundaries sit and how their expertise still matters.

Adoption cannot be solved through access alone. People experiment when they understand the purpose, feel safe to learn and can see how the tool helps with real work. Managers need enough confidence to redesign routines, not simply tell teams to “use AI more.” Governance should be proportionate and practical: clear enough to protect data, quality and trust without making useful experimentation impossible.

The strongest measure of success is not the number of licences or prompts. It is whether work is clearer, decisions are better, capacity is released and people can use that capacity for higher-value contribution. AI earns its place when it solves an organisational problem and produces a better way of working—not when it becomes another layer added to an already crowded process.