← all resources whitepaper

the agentic HR operating model

what changes when intelligence acts inside the process rather than reporting on it, with a maturity model

Most HR analytics tells you what happened. An agent does something about it, inside the process, while the process is still running. That difference changes job design, approval paths, audit trails and the definition of a good HR outcome. Few operating models have caught up. McKinsey found that 23% of organisations were scaling an agentic AI system somewhere in the enterprise, but in any single business function no more than 10% had reached that point.

This paper sets out a five-stage maturity model, from reporting through to autonomous execution under human review. It then maps the impact role by role across HR business partners, recruiters, L&D leads, compensation analysts and shared services teams. For each role it names the tasks that move to agents, the judgement that stays with people, and the new work that appears.

Written for CHROs and HR operating-model owners who have run AI pilots and now need a structure for the stage after the pilot. Includes the authority ladder, a role-by-role impact map, an escalation design, and a 90-day pilot with control groups.

what’s inside

1. what changes when intelligence acts inside the process
2. the five stages of agentic HR maturity
3. the authority ladder: propose, decide, execute
4. a role-by-role impact map for the HR function
5. designing the human checkpoint
6. accountability, audit trails and the record you keep
7. a 90-day pilot and the measures that prove it

what you’ll be able to do

• place your HR function on a five-stage maturity model using evidence rather than ambition
• write an authority ladder that states what an agent may propose, decide and execute for each process
• redraw three HR roles with the tasks that move to agents and the judgement that stays human
• run a 90-day agent pilot with a control group and an agreed measure of decision quality