At the World Economic Forum's 2026 annual meeting in Davos, business leaders reached a consensus that the era of experimenting with AI in workforce planning is over and the era of execution is upon us. The race among organisations is now to redesign operations around it. SHRM's 2026 State of AI in HR report observes that 92% of CHROs expect further AI integration in the workforce, yet 54% of organisations have not implemented any AI in their HR function at all. Bridging the gap between executive expectations and operational reality is the new HR transformation challenge. This demands two shifts: embedding AI in HR operations and replacing job-title-centric talent models with skill-based workforce planning.
What is skill-based workforce transformation?
Skill-based workforce transformation is the shift from organising work around fixed job titles and credential requirements to organising it around defined skills, demonstrated capabilities, and measurable outcomes. In a skill-based model, talent decisions such as hiring, internal mobility, learning, and succession are made against a shared skills taxonomy rather than against role histories and degree filters.
The WEF projects 170 million new roles to emerge by 2030 alongside 92 million job displacements, and estimates that 39% of the core skills required in the current workforce will change within the same period. Jobs designed around stable, predictable tasks are losing relevance. The replacement is work organised around skills, projects, and outcomes that evolve as business priorities shift. About 85% of companies now report using some form of skill-based approaches in talent acquisition.
How is AI changing HR operations and skill-based workforce planning?
AI in HR and skill-based workforce planning are interconnected tracks. AI makes skill-based planning operationally viable at enterprise scale. Together, they are reshaping HR functions and the value those functions create.
AI in HR operations
AI adoption in HR is concentrated where data volumes are highest and structured outputs are most needed. Recruiting leads adoption among organisations that have implemented AI in HR, followed by learning and development. AI automates high-volume screening, generates skill-based job descriptions, delivers structured assessments at scale, and reduces the manual burden of tracking candidate progression. This frees up HR professionals for the judgements that require human insight.
Among large organisations with over 5,000 employees, 60% have already implemented AI in their HR functions. Among organisations with fewer than 100 employees, the figure is 33%. Organisations that have embedded AI into all aspects of the employee lifecycle are operating with faster cycle times, better data, and more consistent decisions than those running similar functions on manual processes, according to SHRM.
Skill-based workforce planning: from static roles to dynamic capability
Static role structures were designed for stability. With changing technologies and economic objectives, skills requirements now change faster than role definitions. Real-time competency tracking empowers businesses to mobilise internal talent dynamically as priorities evolve.
Skill-based workforce planning builds on a common skills taxonomy: a shared language that defines what each capability means, the required proficiency level, and the observable evidence that validates it.
Implementing skill-based hiring and planning in practice
Moving from a traditional model to a skill-based one involves a paradigm shift on five fronts, in this sequence:
Role definition and skills taxonomy
Job descriptions are rewritten to specify skills, proficiency levels, and observable outcomes rather than credentials and years of experience. A shared skills taxonomy standardises what each capability means across roles and teams, preventing each department from defining the same skill differently and ensuring skills data is reusable across hiring, learning, mobility, and planning decisions.
AI integration in screening
Skill-based approaches are most efficient when AI assists with high-volume screening against specific skills criteria, freeing HR personnel for assessment, interviewing, and advisory work.
Assessment redesign
Structured assessments, work samples, simulations, and scenario-based interviews replace unstructured screening. Each skill is scored against a consistent rubric, reducing evaluator bias and producing hiring decisions that are auditable and comparable across candidate pools.
Internal mobility and connected planning
Using the same skills language for hiring and internal movement creates interconnected workforce intelligence. Skills data generated during hiring feeds visibility into existing employee capability, making internal redeployment more deliberate and succession planning more data-driven.
Continuous refinement
Skills definitions evolve with roles and market conditions. Organisations that establish biannual taxonomy governance cycles can reduce skills obsolescence and maintain higher internal redeployment rates.
The data and governance challenge
The primary barrier to AI adoption in HR is limited awareness of its capabilities. There are also technical concerns around transparency, accuracy, and the absence of usable internal data. Skill-based workforce planning amplifies this dependency. A skills taxonomy is only as useful as the quality and consistency of the data mapped against it. Workforce planning on poorly structured or inconsistently defined skills data produces the same imprecision as planning on job titles.
How can Infosys BPM help with human resources outsourcing?
Infosys BPM human resources outsourcing (HRO) services enable enterprises to operationalise AI-driven, skill-based talent architectures at scale. By combining HR consulting, predictive workforce analytics, and global delivery infrastructure, Infosys BPM modernises recruitment, learning, and talent planning across the hire-to-retire lifecycle.
Frequently asked questions
Detailed enough that two managers reading the same skill at the same proficiency level would accept the same evidence, and no more. Taxonomies fail in both directions: too coarse and every role maps to the same handful of generic capabilities; too granular and nobody maintains it. The appropriate number of skills and proficiency levels depends on the organisation’s size, role complexity, and governance capacity.
Not automatically. It means requiring a degree only where the role genuinely depends on what the degree certifies, rather than using it as a default filter. Some roles are regulated or technically dependent on formal qualification. The change is that the requirement has to be justified against the work instead of inherited from the previous job description.
It comes from information the organisation already holds: rewritten job architectures, assessment and interview scores, learning completions, project and staffing histories, and self-assessments validated by managers. Buying an inferred skills library is a starting point, not a substitute; inferred skills that nobody has confirmed will produce workforce plans built on assumptions.
The main ones are bias inherited from historical hiring data, opacity when a rejected candidate asks why, and a growing set of jurisdictional requirements on automated employment decisions, including notice, bias auditing, and human review. Practically, this means keeping a human decision point on rejections, documenting what the model scores, and testing outcomes across groups rather than assuming neutrality.
Look at quality of hire and retention at 12 months, the share of vacancies filled internally, time to fill for hard-to-source skills, and the proportion of candidates progressing from non-traditional backgrounds. Screening speed alone is a process metric; it can improve even when the underlying talent decisions do not.
