Redesign the workflow. Develop the workforce.
Connect process changes to roles, judgement, learning and what happens to freed capacity. Commercial examples and executive accounts sit beside bounded evaluations.
- Redesign the workflow and the workforce together: Use the sponsored workflow-debt example as a design proposition, not an independently demonstrated municipal result.
- Make role owners part of AI workforce redesign: Bring role owners and learning pathways into the redesign before counting agents or cuts.
- DBT: task speed, quality and organisational productivity are different outcomes: Ask why task-level improvements need not become a demonstrated organisational gain.
- Denmark: task changes can precede earnings and hours effects: Keep changed tasks, expectations, hours and earnings as different outcomes.
A decision to discuss: What should stop or change in this workflow, and what do people need to be able to do afterwards?
Borrow a method from an examined AI case.
Compare bounded public-sector cases and return to the actual purpose, information, review work and owner. A sandbox finding is not transferable blanket permission.
- Lidingö and IMY: AI proposes redactions, a person owns disclosure: Follow the boundary between an AI masking suggestion and the official who owns disclosure.
- Talt: map the care workflow before building the AI service: Map the current process and separate a bounded test from operational use.
- Göteborg’s Copilot Chat assessment shows a real governance debate: Read a real municipal discussion of controls, legal caution and operational value without confusing Chat and Premium.
- AI in municipal HR: responsibility starts in the work: Bring the method back to Swedish public-sector guidance and a fresh local assessment.
A decision to discuss: Which conditions can we reuse, and which change when this becomes an HR workflow?
Finished by noon. What happens next?
Follow the gap between the work people really do, the contribution HR recognises and the choices made about released time.
- AI work has changed. Has the job description?: Notice the gap between actual AI work, job descriptions and rewards.
- Invisible work and the right to challenge AI: Include checking, correction and other invisible work in the conversation.
- Denmark: task changes can precede earnings and hours effects: Distinguish task changes from earnings and recorded-hours effects in the revised Danish study.
- Recognise learning as part of the job: Make learning and helping colleagues a recognised part of the job.
A decision to discuss: How should we recognise AI-enabled contribution and decide what happens to the time released?
What should our new AI colleague be allowed to do?
Move from the promise of an agent to a workflow with an owner, a purpose, boundaries and meaningful human judgment.
- Agentic HR: follow the approval boundary into execution: Follow the boundary from a recommendation to an approved action; distinguish supplier claims from tested capability.
- Redesign the work before adding more agents: Examine the process before placing another agent on top of it.
- Agentic AI in practice: early experience needs careful questions: Compare practitioner experience with the vendor proposition.
- AI in municipal HR: responsibility starts in the work: Bring the question back to Swedish public administration and responsibility.
A decision to discuss: Which tasks can an HR agent perform, who owns the result and when must a person decide?
How do people gain capability, not just output?
Connect everyday experimentation to judgment, peer learning and the opportunities people need to keep developing.
- Swedish social partners: learning belongs close to work: Start with the Swedish social partners’ case for learning close to work.
- Public HR needs capability inside the organisation: Distinguish the capability needs of staff, managers and process owners.
- Invisible work and the right to challenge AI: Ask who practises judgment, and who can safely challenge an AI result.
- Me and we with AI: better ideas in a bounded experiment: Compare individual and team work in a specific experiment, without treating it as universal.
A decision to discuss: How can HR protect judgment and build learning into ordinary work with AI?
What makes a municipal pilot become normal work?
Connect the practical HR question to local capacity, shared municipal work, participation and the conditions for sustained adoption.
- Municipal HR can build capability together: Separate local responsibilities from obstacles municipalities can address together.
- A local HR assistant starts with owned guidance: Start with a local assistant built around owned HR guidance and a bounded audience.
- Social dialogue on AI: skills, influence and working conditions: Include employees and their representatives while shaping the work.
- Public-service staffing: scenarios create more choices: Connect the experiment to a real staffing and service question.
A decision to discuss: What must our municipality put in place to turn a useful HR experiment into everyday practice?