Job descriptions have always had a shelf life. AI is making them shorter.

The work sitting inside roles is changing quickly, with tasks that once required hours of human input increasingly supported, accelerated or completed by AI. Yet many businesses are still planning their workforce in much the same way they always have: define a role, write a job description, find a person to fill it.

That approach starts to look increasingly limiting when the work itself is changing faster than the role around it.

Instead, businesses need to start one step earlier. What work needs doing? What capabilities does it require? And only then: who, or what, is best placed to deliver it?

It’s a shift from planning around jobs and headcount to planning around work and capability. And as AI becomes a bigger part of the workforce, it could fundamentally change job architecture.

What is job architecture, and why does AI change it?

Job architecture is essentially the structure businesses use to organise roles, responsibilities, levels and career paths, creating a consistent way to define what different jobs involve and where they sit within an organisation.

The challenge is that this structure assumes the work within those roles stays relatively stable. As AI changes what people can do, and how quickly they can do it, that assumption becomes harder to rely on.

A role might keep the same title while the work underneath it changes significantly, with some tasks automated, others accelerated by AI, and people spending more of their time on the things that still need a human touch, from judgement and communication to problem solving and decision-making.

That doesn’t mean the role disappears. It does mean the job description you wrote a year ago may no longer tell you much about the capability you need tomorrow.

For workforce planning, that’s an important distinction.

Start with the work, not the job

One of the central ideas behind La Fosse’s Beyond Headcount Blueprint is deceptively simple: before deciding who you need, take a proper look at the work in front of you.

That means breaking work down into the outcomes and capabilities needed to deliver it, rather than automatically reaching for a familiar job title every time there’s a gap to fill.

Once you understand the work, the options become much broader. Does it need to stay fully human, could AI take on some of it, or could an AI agent handle it altogether? Would a combination of people and AI deliver a better result, or is there an opportunity to rethink the whole process rather than simply finding a new way to complete the same tasks?

This is where workforce planning starts to move beyond headcount. Instead of automatically asking, “Do we need another five people?”, businesses can ask a much more useful question: “What capacity and capability do we need to deliver this?”

And as AI becomes capable of doing more of the work, the answer to that question is going to keep changing.

People, AI or both? The new workforce planning levers

La Fosse’s AI Workforce Council, who informed the blueprint, discussed a practical example of treating AI as another member of the workforce, with its own capability and cost.

For an individual project, that creates several levers. A business could reduce the scope of what it plans to deliver, increase human capacity, add AI capacity or change the balance between the two.

The point isn’t that AI should replace the person wherever possible. It’s that headcount is no longer the only variable available.

And simply bolting an AI tool onto an existing process isn’t necessarily transformation either.

The Council repeatedly returned to the importance of stepping back and looking at the underlying work. If a process was designed around the limitations of the technology and workforce available at the time, adding AI to one stage may improve it, but it may also miss a much bigger opportunity to redesign the process altogether.

The question therefore isn’t just, “Where can we use AI?”

It’s, “If we were designing this work today, how would we do it?”

What is skills-based hiring?

Once businesses start with the work rather than the job title, hiring changes too.

Skills-based hiring focuses on what someone can do, rather than using previous job titles, qualifications or a conventional career path as proxies for capability.

That’s particularly relevant when roles are evolving quickly. The perfect candidate on paper may have experience built around yesterday’s version of the job, while someone with the curiosity, judgement and ability to learn could be far better equipped for what it becomes next.

AI is also making one of recruitment’s traditional signals increasingly unreliable: the CV.

Anyone can now use AI to produce a polished application. That sparked an interesting debate within the AI Workforce Council.

One view was that if a candidate can use AI effectively to produce an excellent application, that itself demonstrates a capability that businesses increasingly value. Another was that once everyone has access to the same tools, polished applications become less useful for differentiating genuine ability.

There was, however, much more agreement about what comes next.

Test the capability.

Give candidates meaningful tasks, validate what they can do and assess how they think, rather than relying too heavily on how convincingly they describe it on a CV.

What does a skills-based organisation look like?

Skills-based hiring is only part of the picture.

A skills-based organisation understands the capabilities it already has, the capabilities it expects to need and where the gaps sit between the two.

That creates more options than simply recruiting every time something changes.

Some capability can be hired. Some can be developed internally. Some work can be supported or delivered by AI. A specific problem might require an external team rather than another permanent employee. And some gaps will require different leadership altogether.

This is why capability gap analysis becomes increasingly important as AI adoption accelerates.

If you don’t know what capability you already have, it’s difficult to make a sensible decision about what you need next.

Building a workforce around capability

There isn’t one answer to every workforce gap, which is precisely the point.

La Fosse works across different parts of that decision. La Fosse Recruitment helps organisations hire around capability, curiosity and judgement. La Fosse Executive finds leaders who can set direction and accountability. La Fosse Academy develops skills from within, while Inovus by La Fosse provides complete teams and solutions where the requirement is an outcome rather than another individual hire.

The starting point is the same in each case: understand the work and the capability it requires before deciding how to fill the gap.

Because as AI changes what work looks like, building a workforce around static job descriptions becomes increasingly difficult.

The job title may stay the same.

The job probably won’t.

Beyond Headcount explores how businesses can move from traditional headcount planning towards a workforce strategy built around capacity, capability and the work that needs doing.

Download the Beyond Headcount Blueprint – coming soon