Because if AI is changing roles, skills, work and ultimately the shape of the workforce, somebody needs to help redesign all of it. And CHROs can’t credibly redesign everyone else’s role without rethinking their own first.
The assumption the people function was built on
For most of its history, the people function has been built around a fairly safe assumption: most of the capability a company needs comes from the people it employs. You hire them, develop them, organise them, reward them and retain them. The workforce was largely human, largely employed and conveniently visible on an organisational chart.
That model is changing underneath them. Companies can increasingly develop capability internally, hire it permanently, access specialist expertise when they need it and deploy AI alongside people. Jobs themselves are becoming less fixed as machines take on some activities, augment others and increase the value of work requiring human judgement.
Which means the CHRO’s remit has to move with the thing they’re being asked to redesign: from responsibility for the people a business employs towards responsibility for the capability the business needs to perform.
McKinsey describes something similar as HR’s “dual mandate”: transforming the people function while helping redesign the wider organisation around AI. Their research highlights the gap between adoption and impact too. While 88% of companies use AI in at least one function, only 39% report a material contribution to EBIT. Plenty of AI, considerably less transformation.
CHROs can’t credibly redesign everyone else’s role without rethinking their own first.
Understand the work before you approve the hire
For the mid-sized businesses we work with, the challenge is usually rather less theoretical. A capability gap appears on Monday and somebody wants permission to hire someone by Friday. Increasingly, there needs to be a step in between: understand the work. What actually needs doing? What can AI absorb? Where does human judgement matter? What’s still missing once you’ve answered those questions? Only then do you know whether you actually have a hiring problem.
For a long time, the job has been the basic unit of workforce planning. But a job is really a collection of activities, and AI doesn’t affect each equally. Some can be automated, some dramatically accelerated and others become more valuable because they require judgement, accountability or human interaction. Two businesses can now employ the same number of people and have radically different capacity depending on the quality of their talent, the intelligence surrounding them and how effectively the work has been designed.
Workforce planning starts to look like capital allocation
Once that’s true, workforce planning starts looking suspiciously like capital allocation. What capability should we own? What should we access? Where is exceptional human judgement worth paying for? What should machines handle? And where will the next dollar of workforce investment generate the greatest return?
For years, HR has talked about becoming more commercial to protect its seat at the table, which I’ve always found slightly odd. Being commercial isn’t simply understanding the P&L, controlling headcount or defending the workforce budget. It’s helping determine how one of the largest investments in the business should be deployed in the first place.
Businesses are remarkably good at creating complexity and then employing people to manage it.
Sometimes it isn’t a workforce problem at all
Sometimes the workforce cost problem isn’t actually a workforce problem. It’s an organisation design problem we’ve been hiring our way around. Businesses are remarkably good at creating complexity and then employing people to manage it. More layers create more coordination, more coordination creates more meetings, and eventually more people get hired to navigate the complexity the organisation created itself. Then Finance looks at the wage bill and announces a productivity problem.
AI gives us an opportunity to start somewhere else. Most businesses have a relatively small number of decisions where better human judgement has an outsized impact on performance. AI can synthesise information no individual could reasonably process, model possibilities and challenge assumptions, giving those people considerably more intelligence to work with.
Which makes one of the more useful workforce questions in the AI age: how much human judgement does this business actually need, and where does it create the most value?
Design from the work backwards
Start there and the organisation begins to look different. Instead of designing from the org chart forwards, design from the work backwards. Understand the outcomes and decisions that matter, determine where humans create disproportionate value, what machines can do better, then build the capability around it.
Increasingly, that capability won’t all sit on the payroll. Businesses can have a strong internal core holding context, culture, relationships and judgement, surrounded by specialist expertise, partners, flexible talent, offshore capability and AI that flex with what the business actually needs. Less fixed capability, more available capability.
And that changes the CHRO’s job. They’re no longer simply leading a People team. They’re increasingly orchestrating an ecosystem of capability, and they can’t do that from inside P&C. It means working hand-in-hand with the rest of the C-suite to rethink workflows, determine where human judgement matters, decide what technology should handle and redesign roles around the result.
They’re no longer simply leading a People team. They’re increasingly orchestrating an ecosystem of capability.
Aiming at too small a target
Which is why putting AI into recruitment, onboarding, learning and employee service, while useful, is aiming at too small a target. If AI can finally kill a few HR spreadsheets nobody reads, it deserves our gratitude. But that’s AI transforming HR. The bigger opportunity is for P&C to help AI transform the business.
P&C has spent years trying to become more commercial, and AI may have just handed it the opportunity. But that expanded remit isn’t automatic. CHROs will need to earn it by helping the rest of the C-suite solve the problems they already care about: capability, capacity, productivity, cost and performance.
The opportunity isn’t to own another part of the business. It’s to become a more consequential partner in designing and orchestrating its capability: how work gets done, where capability comes from, where human judgement creates value and how people and machines fit together.
Which is why the CHRO’s biggest opportunity isn’t really in HR anymore. It’s helping redesign the business around what people and machines can now do together.
Sources
- AI in HR transformation: A dual mandate — McKinsey & Company. https://www.mckinsey.com/capabilities/people-and-organizational-performance/our-insights/the-organization-blog/hrs-dual-mandate-in-the-ai-era
- The state of AI in 2025: Agents, innovation, and transformation — McKinsey & Company. https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai