A new title arrived in the C-suite conversation this week, and it is a good one. Making the case in the national press, the chief executive of the Chartered Management Institute argued for a Chief Work Officer — a single executive whose remit is the work itself: how it is designed, who does it, which parts move to machines, and how fast the whole arrangement can change.[^1] The reasoning is that the old division of labour has broken. Technology strategy and talent strategy can no longer sit in separate offices, because the question that now matters most — what is the work, and who or what should do it — sits across both. A Chief Work Officer, the argument goes, retires the static job description, organises around skills rather than the org chart, and plans in quarters rather than decades.
We think the diagnosis is right. We also think the title, on its own, will not be enough — and the reason is the whole of this week's other story.
Because in the same week, the market gave a sharp demonstration of why. The frontier-model access that a government directive cut a fortnight ago began to return — but not as it left. It returned stratified: restored in tiers, mediated by government, scoped to trusted partners and jurisdictions rather than switched back on for everyone at once.[^2] Read those two stories together and the shape is unmistakable. The work is being redesigned around an intelligence layer whose availability is now a tiered, politically-mediated privilege. A title that owns "the work" inherits that exposure on day one. Whether the title can do anything about it depends entirely on what sits behind it.
The role the market just named
The case for the role is strong, and we will not weaken it by overstating it. When the work changes continuously, an operating model built to change occasionally is a liability, and somebody senior has to own the rate of change itself. Splitting that ownership across a technology office that buys the tools and a people office that manages the humans guarantees the seam where most transformation fails — the gap between the capability that was purchased and the work that was supposed to change. Naming a single owner for that seam is a genuinely useful move.
The supporting claims that travel with the proposal — that smaller organisations out-adapt larger ones because they carry less structural drag, that the usable life of a given skill is collapsing — are the arguments being advanced in that conversation, and they are plausible. We carry them as the author's cited research and argument, not as independently verified figures.[^1] The directional point stands without them. The work is no longer a fixed thing that a person is hired to do; it is a moving thing that has to be continuously re-cut between people and machines. That requires an owner. The proposal is to give that owner a title.
A title is not a discipline
Here is the part the enthusiasm tends to skip. We have watched the C-suite name new roles before in response to a structural shift, and we have watched a predictable share of them fade. The Chief Digital Officer arrived to own "digital," and in many organisations dissolved back into the business the moment digital stopped being a separate thing. The pattern is not failure of intent. It is that a title creates an owner without, by itself, creating the capability the owner is now accountable for. The nameplate goes on the door on Monday. The method the office is supposed to run does not arrive with it.
A Chief Work Officer is, precisely, an accountability looking for a discipline. The mandate is clear enough — own the work, own the rate at which it changes. The harder question is the operating one: by what method? When the board asks the new officer to show how the organisation will absorb continuous change rather than merely survive it, what is the answer made of? That is not a question a title answers. It is a question a discipline answers — and the discipline has, until recently, had no name.
What the mandate actually requires
Strip the role down to what it must actually do, and five obligations fall out of it.
It must absorb continuous change — convert a constant stream of new tools, models, and methods into working practice faster than the change arrives, rather than accumulating a backlog of half-adopted capability that quietly decays.
It must govern transformation — decide which changes are pursued, in what order, against what evidence, with the authority to stop the ones that are not earning their disruption.
It must redesign work — break roles down into capabilities and recompose them around what humans and machines each do best. That is the operational meaning of "skills over the org chart," and it is considerably harder than drawing a new chart.
It must own capability rather than rent it — keep the parts of the operating system that constitute the organisation's actual advantage inside the organisation, rather than leasing them, by default, from whichever vendor was most convenient at the time.
And it must convert technology into measurable operating advantage — close the distance between what has been bought and what is actually moving in the numbers, which is the exact place most transformation budgets quietly die.
Those five obligations are not a job description. They are a discipline, and it is the one this series has spent its run naming: Transformation Capability Management. TCM is the client-side discipline that makes a Chief Work Officer's mandate executable — the method that turns "own the work" from an accountability on an org chart into a capability the organisation actually holds. The title is the role. TCM is the discipline behind it. Without the second, the first is a nameplate.
Why ownership is the load-bearing one
Of the five obligations, one is load-bearing this week, and it is ownership. The other four assume a stable supply of the technology being absorbed, governed, designed around, and converted. This fortnight removed that assumption in public.
When frontier-model access returns in tiers — scoped by jurisdiction and partner status, restored to some and withheld from others — the organisation that redesigned its most important work around a single rented intelligence layer has redesigned its work around a privilege it does not control. The Chief Work Officer who owns the capability — the context, the routing, the review gates, the decision rights, the standing ownership of the work regardless of which model is available this quarter — keeps the operating system when the substrate is rationed. The Chief Work Officer who rented the implementation layer finds out, mid-restoration, that the layer was never theirs to keep.
This is the divide the analysts have started to gesture at — the widening gap, framed at a recent industry symposium, between the organisations turning AI access into compounding operating advantage and the ones whose access never reaches the numbers.[^3] After this fortnight, the dividing line is not who has the better model; nobody is guaranteed the better model now. The dividing line is who owns the capability around it. A Chief Work Officer whose mandate includes ownership is building something that survives a tiered restoration. A Chief Work Officer whose mandate stops at adoption is building on rented ground.
What this asks of you
If your organisation is considering a Chief Work Officer — or has quietly already created the role under a different name — the title is the easy part. The discipline is the work. Appoint the officer, by all means. But resource the officer with a method for absorbing change, governing transformation, redesigning work, owning capability, and converting technology into advantage. A nameplate on an office door is not an operating system, and the gap between the two is exactly where the last decade of new C-suite titles went to fade.
The full cross-layer reading of this week — the stratified restoration, the new role, and the discipline the role requires — sits inside The Capability Brief — Edition 14 and on the intelligence page. The week's enterprise-software detail is in SAP Compass — Edition 8, and the signal summary is in Signal — Edition 18.
Your Waypoint
One mandate. One owner. One test.
Take the five obligations above — absorb, govern, redesign, own, convert — and, before you appoint anyone, score your organisation honestly against the fourth one: ownership. Pick the single most important piece of work your organisation has redesigned around AI in the last year. Ask one question of it. If the model behind it were rationed tomorrow — restored only to a tier you are not in — what would still be yours? The context, the runbooks, the routing, the review gates, the decision rights, the person who owns the work regardless of which model answers the phone this quarter?
If the honest answer is "most of it is ours," you have a capability, and a Chief Work Officer would have something real to manage. If the honest answer is "we are not sure," you have just found the first item on the new officer's desk — and the reason the title, by itself, was never going to be enough. The role is the easy part. The discipline behind it is the job.
References
[^1]: Ann Francke (Chief Executive, Chartered Management Institute), "Could a chief work officer help us harness AI better?", The Times, 24 June 2026 — the role "should bridge the divide between tech and talent, turning a rigid hierarchy into a responsive network"; technology and talent strategy can no longer sit in separate offices; "the static job description is dead" in favour of a skills-based approach; "the rigid ten-year plan is gone… your execution must be updated quarterly." Primary-verified against the article text, 4 July 2026. Research claims advanced in the piece — including CMI research showing smaller organisations out-adapting larger ones, and the accelerating turnover of task allocations — are carried here as the author's cited research and argument, not as independently verified figures.
[^2]: Frontier-model access (Anthropic Mythos 5 / Fable 5), suspended by a U.S. export-control directive on 12 June 2026, began a tiered, government-mediated restoration over the following fortnight; as reported around 26 June 2026, access returned stratified — scoped by jurisdiction and trusted-partner status rather than restored universally at once. Carried from the prior Way Pointer cycle and corroborating the "structured access for trusted democratic partners" framing advanced at the G7 AI lunch, 17 June 2026. Reported; restoration terms remain in flux at time of writing.
[^3]: "AI gains divide" — described here using a directional framing surfaced at a recent industry symposium (the widening gap between organisations converting AI access into measured operating advantage and those whose access never reaches the numbers). Used as a framing only, not a proprietary analyst term, and not attached to a specific quantified projection.
[^4]: BusinessGPS, Transformation Capability Management v2.2, 21 May 2026; prior Way Pointer editions — The Off-Switch Has a Price Now (Edition 17, 22 June 2026) and When the Substrate Gets Switched Off (Edition 16, 15 June 2026); The Capability Brief — Edition 14 and SAP Compass — Edition 8, 29 June 2026 (companion intelligence editions).
The Way Pointer is published by BusinessGPS — the transformation capability platform helping organisations make strategic choices about what they build, what they buy, and what they own.