Rebuilding the Organisations for the Age of AI: Work, Workforce, Workplace

Andrea Kahudova

Aug 25, 2026

How to rebuild an organisation for a future nobody can predict

Rebuilding the Organization for the Age of AI: Work, Workforce, Workplace

Three-quarters of today's jobs will need redesign, upskilling or redeployment by 2030. Here is how to rebuild an organisation for a future nobody can predict – and where to find the insight to do it.

Every transformation I have led in my career had one thing in common: we were rebuilding the organization for a future we could describe. A merger, a new market, a new system – the destination was known; the work was getting there. The AI transformation is different in kind, not just in size. Nobody – not the vendors, not the analysts, not the people building the models – can tell you what work looks like in 2030. And yet standing still is not a neutral choice: McKinsey estimates that three-quarters of current jobs will require redesign, upskilling or redeployment by then. The organizations that wait for clarity will be redesigned by events instead.

So how do you rebuild an organization for a future you cannot see? You stop designing for a destination and start designing for adaptability. In practice, that work has three fronts – the work, the workforce, and the workplace.

Front one: the Work – redesign the process before you automate it

The most expensive mistake of this decade is already visible: companies automating their existing processes exactly as they are. AI makes a bad process faster – which mostly means you produce the same waste at higher speed. The old rule of every digitization wave applies with double force now: review and clean the processes first.

The discipline that works is task-level redesign. Decompose each process and role family into its actual tasks – not the job-description fiction, the weekly reality. Then ask three questions of every task, in this order: Should this exist at all? (Many reports, approvals and handoffs exist because they always have.) Should it be redesigned? (The process built around paper rarely suits a world of instant drafts.) And only then: should it be automated, augmented, or kept human-core? Automation is the third question, not the first – that ordering is the whole craft.

▪ What this means practically: start with your most painful end-to-end process, not your org chart. Map its tasks, kill the no longer needed steps, redesign the flow around what AI does well (drafting, summarizing, routing, answering) and what humans do irreplaceably (judgment, exceptions, relationships, accountability). The org chart follows the redesigned work – never the reverse.

Front two: the Workforce – human and agentic

The workforce of the future is not smaller. It is different in composition: human and agentic. AI agents – software that doesn't just answer but executes multi-step work – are joining teams as digital workers: they check the system, draft the response, file the document, escalate the exception. Deloitte calls it the silicon- based workforce; Microsoft's Work Trend Index describes 'frontier firms' already managing blended human- agent teams. Whatever the label, the management questions are refreshingly familiar HR questions: who supervises the agents, who is accountable for their output, how is quality reviewed, how do humans and agents hand work to each other – and what happens to the people whose tasks the agents absorbed?

That last question is where redesign becomes leadership. The honest answers are upskilling and reskilling at a scale most companies have never attempted – and the economics favor courage: redeploying displaced capacity into adjacent roles almost always costs less than redundancy-plus-rehire, and it keeps the institutional knowledge you already paid for. Two design notes matter. First, the rise of the generalist: as agents absorb

specialist routine, broader, outcome-owning roles grow – design career paths for range, not just depth. Second, a finding worth pinning to the wall: organizational factors – culture, manager support, talent practices – explain roughly twice as much of AI's real impact as individual skills and mindset. You cannot train your way out of a design problem.

Front three: the Workplace – the operating system of work

The third front is the environment where human and agentic work meets: the workplace as an operating system. That means the digital estate (which tools, which data, which agents are allowed to touch what), the governance (your AI policy, the EU AI Act guardrails, the works-council agreements), the management rhythm (how hybrid human-agent teams are led, reviewed and developed), and the culture (whether using AI is disclosed or hidden, whether learning is rewarded or quietly punished). A redesigned process staffed by a redesigned workforce still fails inside an un-redesigned workplace – this is where most pilots go to die.

Designing for a future you can't see

If the destination is unknowable, adaptability becomes the design goal. Five principles carry it:

▪ Sense on a rhythm, not on a fright. An annual task scan – what became automatable this year? – plus a quarterly

look at what your own tools quietly switched on.

▪ Make small, reversible bets. Pilot one process end-to-end, measure, then scale or stop. Options are cheaper

than plans; the portfolio of experiments IS the strategy.

▪ Build learning velocity as infrastructure. With technical skills carrying a roughly four-year half-life, the speed at which your organization acquires the NEXT skill is the only durable advantage. Protect learning time structurally; measure time-to-competence, not training hours.

▪ Keep the human spine explicit. Judgment, relationships and accountability stay human – say so in writing. People absorb enormous change when they know what is stable.

▪ Change management is the multiplier. Redesign without adoption is a slide deck. Communicate early, involve works councils before you must, protect the clumsy first attempts – the classics, because they compound.

Where to get insight and data

You don't have to navigate blind. Externally, a small trusted set beats a flood: the World Economic Forum Future of Jobs Report for the macro numbers, McKinsey and Deloitte for the agentic-workforce research, Microsoft's Work Trend Index for what's actually happening inside companies, Josh Bersin's work for the HR-specific frame, and the OECD for the policy horizon. Follow what competitors ship, not what they announce. Internally – and this is the underused goldmine – your own organization is the best dataset you own: task decompositions, skills inventories, pilot metrics, hours returned, adoption curves. Your own experiments are the only data that is certainly true for you.

Whose job is this?

Organizational redesign in the AI age is not an IT program with an HR appendix. It is org design, workforce economics, capability building and change leadership – which is Leadership & HR's job description at its most strategic level. Both are needed to map the tasks, plan the reskilling, design the roles, negotiate with the works council and carry the culture – as this collaboration decides whether AI becomes a productivity success story or a trust and change crisis.

The future of your organisation isn't something you predict. It's something you design – one process, one role, one skill at a time. And the best time to start designing is while the future is still negotiable.

– Andrea Kahudová, FutureLeads. I help organizations redesign work, workforce and workplace for the age of AI – practically, safely, and with people at the center. futureleads.eu

Sources & data points

▪ McKinsey: ~75% of current jobs require redesign, upskilling or redeployment by 2030; 'The future of work is agentic'.

▪ MIT Technology Review (2026): Rethinking organizational design in the age of agentic AI; PwC: 'No more pyramids'.

▪ Microsoft Work Trend Index 2026: frontier firms, human-agent teams; Deloitte Tech Trends 2026: the silicon-based workforce.

▪ Research 2026: organizational factors explain ~67% of AI's realized impact vs ~32% for individual factors.▪ WEF Future of Jobs 2025: 170M/92M/+78M by 2030; ~40% of core skills change; ~4-year technical skill half-life.