The actor grants the system authority with explicit edges — how much it may spend, touch, or decide, and for how long — so delegation can start small and grow with demonstrated competence.
Forms
Recurring forms: autonomy levels per task (suggest / confirm / execute), budgets in time, actions, or money, per-capability rather than per-application grants, and thresholds that automatically demote autonomy when risk rises.
Each of these is stated as a clause, and past the simplest grant the actor is composing several of them at once — which makes the editing surface rule composition, with the same problem of reading back what the bounds will actually permit.
Consequences
- granted authority now has explicit edges — scope, spend, time, risk class — so trust can be extended in increments instead of all at once
- crossing a bound is a designed event rather than a failure: the work pauses and control returns to the actor for renewal, escalation, or takeback
- the bounds only bind if the actor understood them at grant time; authority granted through an unread scope screen is consent in name only
Resources & references
- AI UX Playground — Autonomy Budgets, Per-Action Autonomy, Escalation Thresholds
- aiuxdesign.guide — Autonomy Spectrum, Mixed-Initiative Control
- AI Interaction Atlas — Delegate Authority
Appears in
Governing autonomy
- Setting the limits, step 1 — Set limits on what the agent may do
- Adjusting the limits over time, step 2 — Adjust the limits based on the record
Related patterns
Precedes
- Handoff — crossing a bound is a designed event rather than a failure: the work pauses and control returns to the actor for renewal, escalation, or takeback
- Approval gate — crossing a bound is a designed event rather than a failure: the work pauses and control returns to the actor for renewal, escalation, or takeback
Serves
- Delegation — bounded autonomy covers delegation's set-up station — the shape of the authority granted before the agent runs
Enacts
- Agency — autonomy becomes something the actor meters out rather than switches on
Complements
- Cost transparency — a budget only binds when spend is visible against it
- Capability & scope — what the system can do versus what it may do on this occasion
- Rule composition — scope, spend and time bounds are composed the same way — clauses on what the system may do, rather than on when it acts
Related to
- Settings — where standing bounds tend to live between grants