You have been asked to take cost out with AI. Do you know what the business would lose with each role?

Most plans cover the numbers and the communication. Fewer cover the knowledge, and that is the part that is hardest to get back.

THE SHORT ANSWER
Lead in this order: define the outcome, find the knowledge each role holds, name who could cover it, capture what matters before any role changes, then redesign how the work runs. Cutting headcount alone does not do the last step.

What does it mean to lead through AI-driven change?

Leading through change, sometimes called change leadership, means steering an organisation through a shift in how work is done, and bringing people with you. When AI is changing roles, some will be reshaped, some will move and some may go.

Most guidance on AI change management covers the plan and the message. The part leaders can miss is deciding what the organisation must keep, because the answer is rarely on an org chart. Here, core knowledge means what the organisation depends on that experienced people hold and that is not written down.

Why do cuts that look sensible lose core knowledge?

AI is not the only cause of current cuts. Bloor’s correction article, which focuses on the technology sector, describes a cyclical correction after over-hiring in 2020 and 2021 alongside a longer structural change, and argues that attributing the cuts to AI alone is incomplete. Whatever the cause, the knowledge question is the same. When the headlines call it AI redundancy or AI layoffs, the risk to what the business knows applies just as much.

The knowledge is not on the chart. Org charts show roles and reporting lines. They do not show who knows why a process works, which exception matters or who the customer trusts.

It leaves before anyone looks. Once a role is removed or the person has gone, the chance to ask is gone with them. A handover note rarely carries it.Systems may not make up for it. Bloor argues that the knowledge carried in those roles is being “extracted into digital systems or lost entirely”. That is its argument, not a measured finding. It also argues that removing headcount without redesigning the operating model reduces the immediate cost but does not solve the underlying structural problem.

What does the UK data say?

A minority of employers expect cuts. In the CIPD’s Autumn 2025 Labour Market Outlook, 17% of employers expected AI to shrink their workforce over the next year. The base is set out in our post on strategic workforce planning.

Few have a plan for who steps in. In Robert Walters research reported by People Management in May 2026, 14% of UK organisations had a succession plan in place.

The two surveys measure different things and cannot be added together. For a leader, they point the same way: expect change, and know who could cover the roles that matter.

How should leaders sequence AI-driven change?

  1. Start with the outcome, not the cost target. Decide what the organisation needs to deliver, then plan the work and the workforce behind it. Strategic workforce planning is the place to do it.
  2. Map the work and the knowledge before the roles. A skills gap analysis shows what people can do. Also ask what each role holds that is not written down, which is often called tacit knowledge.
  3. Name who could cover the roles that matter. Succession planning works best when it covers the knowledge the successor needs, not only the title.
  4. Capture before you change anything. Talk through real decisions with the people who hold the knowledge, and record the how and why in a living knowledge system: one built around real decisions, kept in use and owned by someone. Exit interviews and handovers alone will not carry it.
  5. Redesign how the work runs, then decide the structure. Removing roles without changing the operating model leaves the structure that created them. Organisational restructuring should follow the redesign, not replace it.

Decisions that affect people’s jobs carry legal obligations. This post does not cover them, so take specialist employment advice on the process.

Where does Bloor’s thinking fit?

Bloor’s OAI℠ (Original Augmented Intelligence) research keeps the human at the centre, so the leader’s first question is what the people in the organisation know, not only what a tool could replace. Bloor’s Hire to Heir describes a stage in which “nobody owned the intelligence architecture of the augmented organisation”. For a board, it is a useful question to put on the agenda: who owns what the business knows?

Bloor Research helps organisations understand where they are now, where they need to be and what is happening inside the business, including through its X-RAY diagnostic. Its FusionWork™ category looks at leading and planning a workforce of human and digital workers. If you are leading your organisation through AI-driven change, contact Bloor Research to talk it through.

Frequently Asked Questions (FAQs) 

1. What does it mean to lead through change?

Leading through change means steering an organisation through a shift in how work is done, and bringing people with you. When AI is changing roles, it also means deciding what the organisation must keep as the work changes.

2. What should leaders do first when AI changes roles?

Start with the outcomes the organisation needs, then map the work and the knowledge behind it before deciding anything about individual roles. Find out what each role holds that is not written down, and who could cover it, before any role changes.

3. What is core knowledge?

Core knowledge is what the organisation depends on that experienced people hold and that is not written down: judgement, history, exceptions and relationships. It is often called tacit knowledge, and it does not appear in job descriptions or on an org chart.

4. How do you keep core knowledge when roles change?

Identify who holds it, ask about decisions rather than tasks, let successors work through real cases alongside the people who hold it, and record the how and why somewhere people will use. Do this before roles change, not after. Decisions that affect people’s jobs carry legal obligations, so take specialist employment advice on the process.

5. Is cutting headcount enough when AI changes the work?

Bloor Research argues that removing headcount without redesigning the operating model reduces the immediate cost but does not solve the underlying structural problem. It also argues, as an argument rather than a measured finding, that the knowledge carried in cut roles is being extracted into digital systems or lost entirely. It treats AI as one factor among several in current cuts.

6. Who should own knowledge in an organisation that uses AI?

There is no settled answer. Bloor’s Hire to Heir describes a stage in which “nobody owned the intelligence architecture of the augmented organisation”. A practical starting point is to give named leaders responsibility for identifying and keeping the knowledge each critical role holds.

Sources

  1. CIPD, One in six employers say AI will shrink headcount: Autumn Labour Market Outlook, press release 10 November 2025. Total sample 2,019 senior HR professionals and decision-makers in the UK; fieldwork 19 September to 14 October 2025. The 17% figure excludes employers who did not know whether AI tools are used by employees in their organisation (6% of the total sample).
  2. People Management, Only one in seven organisations have a leadership succession strategy in place, research finds, 6 May 2026, reporting research by Robert Walters. 39% of the 250 organisations surveyed (a mix of large and SME businesses) had no succession plan; the article does not give fieldwork dates.

Print Article for this blog Download here