Toby HoughFor years, HR has talked about moving towards skills-based work. The logic is compelling as skills evolve faster than job titles and offer organisations far greater flexibility in how work gets done.
Despite the discussion, progress has been slow. Many organisations remain anchored to role-based structures, not because they believe they are fit for the future, but because skills have proven tricky to operationalise at scale.
The World Economic Forum estimates that 44% of workers’ skills will be disrupted by 2027. That level of change cannot be managed through periodic role redesigns or reliance on external hiring alone. Organisations need a clearer, more dynamic view of the skills they have, and the skills they are about to need. Getting there will require a discussion on skills fragmentation.
The skills challenge is a data challenge
The biggest obstacle to skills-based work is a lack of a shared, usable skills infrastructure.
There is no universal skills language used consistently across education providers, employers and individuals. Training syllabi describe capabilities one way, role descriptions another, and CVs yet another. Even internally, teams doing similar work often use different terminology to describe the same skills.
As a result, skills data quickly becomes unmanageable, and frameworks grow unwieldy. Definitions drift and confidence in the data drops. When skills are hard to trust, organisations revert to what feels like more of a known entity: job titles, qualifications and years of experience.
This is why so many skills initiatives stall. Not because the concept is flawed, but because the system cannot sustain it.
AI changes what’s possible
At scale, artifcial intelligence (AI) has reshaped what it means to be skills-based. It’s also given us a clear focus: helping our organisations and ourselves build the skills and confidence to be truly AI fluent. Manual approaches cannot keep pace with the volume, velocity and variation of skills data across a modern organisation.
AI makes skills usable by doing what humans cannot efficiently do alone – building and maintaining large skill catalogues, cross-referencing inconsistent language, not to mention continuously updating skill profiles as work changes.
Just as AI-powered recruitment tools already bridge variation in job titles, AI can connect different descriptions of capability across role profiles, learning content and experience records. A training syllabus can be analysed against an internal skill framework, or a skills-based CV can be interpreted in context. Similar capabilities can be recognised even when the wording differs.
This is not about creating a perfect or static taxonomy, but making skills visible, searchable and comparable enough to inform decisions.
Skills don’t fail; disconnected systems do
Technology alone is not the solution. Treating skills as a standalone HR initiative is a common mistake.
For skills to genuinely replace job structures as the organising logic of work, learning systems, workforce planning, talent marketplaces and performance processes must be connected. Skills data needs to flow between them, not sit in isolation.
Without that connectivity, organisations risk recreating traditional structures with new labels. Skills frameworks exist, but hiring, progression and work allocation still default to roles because they are simpler to interpret. Over time, confidence erodes and the organisation slides back to familiar known entities (like job titles) for capability.
From static frameworks to living systems
When powered by AI and embedded across systems, skills become dynamic. Gaps are identified earlier, and adjacent and transferable skills surface more easily. Learning aligns directly to real work opportunities rather than generic development plans.
This also changes the employee experience as people gain clearer visibility into how their skills are recognised, how they can grow, and where opportunity exists internally. That transparency is essential if organisations want skills-based models to support mobility and retention, not just planning.
Importantly, and like with all changes, this shift requires trust. Sharing skills data, learning records and experience profiles demands openness and consistency.
What HR should focus on now
The next phase of skills-based work is not about refining frameworks but making skills easier to use than job titles.
That means investing in AI capabilities that reconcile language across systems, designing processes that embed skills into everyday decisions, and enabling managers to use skills data in practical, human conversations about work and progression.
The organisations that succeed will be the ones that make skills visible, trusted and actionable at speed and, in doing so, build a workforce that can adapt as quickly as the work itself.
Toby Hough is VP of people and culture at HiBob
