Workforce planning is for most organizations a quiet HR topic: capacity forecasting, headcount budgeting, vacancy pipelines. Until you end up in a restructuring. Then the question sharpens: which teams have we too many people, which roles will be redundant soon, which workers will likely disappear first? If AI helps determine which side of that line someone lands on, you are in perhaps the heaviest 4(b) area of the EU AI Act.
This post explains where AI sits in workforce planning, why restructuring context produces the heaviest legal stack, and what HR directors must arrange before every new planning cycle.
Where AI in workforce planning sits
The modern workforce planning stack increasingly covers:
- Headcount forecasting AI — Visier, Workday Adaptive Planning, Anaplan: predicting future capacity needs based on growth scenarios and attrition models
- Skill gap analyses at organization level — Talent Intelligence Hub-like systems aggregate skills versus future-state requirements
- Restructuring scenario tools — AI suggests reorganization alternatives, simulates headcount effects
- Workforce optimization AI — proposals for team structure, span of control, layer reduction
- Attrition prediction at individual level — focused on retention actions, but can also be input for redundancy decisions
- Performance + potential matrices with AI scoring — 9-box grids where AI helps determine positioning
- Compensation cost optimization — AI proposals where pay reduction or restructuring is possible
In quiet times this looks like supporting work. In restructuring context the same tools become the heart of life-changing decisions.
Why workforce planning produces the heaviest stacking
Three legal frameworks stack on workforce planning AI in restructuring:
- EU AI Act Annex III point 4(b) — decisions about work-related conditions, task allocation, contract termination
- GDPR — processing of personal data for decisions about workers
- Labor law and works council representation — advisory or approval rights for works council on major organizational decisions and on systems affecting workers
In the Netherlands additionally:
- WOR article 25 advisory rights for major decisions (including restructuring)
- WOR article 27 approval rights for HR policy arrangements
- CAO provisions around redundancy schemes
- UWV test for redundancy on economic grounds
For an employer deploying AI to develop restructuring scenarios: each of these legal layers can be subject to transparency requirements. It is no longer sufficient to say "we have analytically substantiated". You must show which AI role fed which decision, with which validation and which oversight.
When is workforce planning AI within 4(b)
- Headcount forecasting at aggregate level — usually outside 4(b). Market- and business-driven scenarios without individual decisions.
- Skill gap analyses for organizational strategy — usually outside 4(b). Strategic input.
- AI for team restructuring suggestions with names — within 4(b). Person-specific outcomes.
- Attrition prediction at individual level — within 4(b), especially if output feeds retention or redundancy decisions.
- 9-box performance/potential matrices with AI scoring — within 4(b) for individual positioning.
- Workforce optimization tools that designate individual roles for reduction — within 4(b), and in restructuring context under additional obligations (works council, dismissal law).
The line runs along the person level. Aggregate planning is usually lighter; once AI designates individual workers (for retention, training, mobility or redundancy), you are in 4(b).
Step-by-step for workforce planning AI dossier
Document aggregate versus person-specific explicitly
The difference between "our headcount model predicts 50 fewer roles in 2027" and "AI has identified these 50 people" is legally enormous. Document per tool which way.
Build works council track before planning cycle
Waiting until restructuring time to engage works council is practically and legally unwise. Plan AI-related advisory request at planning cycle, even when no restructuring is on the table.
Build scenario archiving for later audit
At future restructuring you may be asked what AI said earlier. Archive scenario output with date, input assumptions and which decisions followed.
Test your AI against the Article 6(3) filter
Interactive self-assessment, updated for the Commission guidelines of 19 May 2026. 9 steps, personal report with reasoning, vendor questions and next steps.
Frequently asked questions about workforce planning AI and the AI Act
Practical questions for HR directors on Annex III point 4(b) in planning and restructuring context.
What to do now
For HR directors with workforce planning AI (and in 2026 that is most): treat this as priority-1 within your 4(b) trajectory. Tool inventory this month, works council conversation about AI in planning well before the next planning cycle, FRIA with restructuring scenario impact within 60 days. Document via the HR AI hub and the HR AI Evidence Pack.
For related 4(b) topics: see the posts on performance reviews, worker monitoring and compensation.