Instead of binary "ask everything" or "do everything," users grant per-tool and per-action autonomy: which actions the agent takes freely, which need approval, which are forbidden. The 2026 reality is per-tool HITL gates more than continuous sliders, but the underlying move is the same — calibrated trust by category, durable across sessions.
Salesforce Agentforce · Cursor Agent Settings · Claude Code Permission Modes · n8n HITL per-tool gating · Tesla Autopilot
Concept Frame
Control is a spectrum, not a switch.
Copilot, delegation, autonomous — three points on a continuum users move along as trust accumulates. The dial in the demo below makes that continuum operable per task and per category.
From the Essay
Control isn’t binary. It shifts with task, experience, and risk.
This prototype already contains the contrast between chat and the pattern response. It stands as the primary comparison surface for this page.
Loading Demo
Preparing the interactive example for this pattern.
✦
One toggle for the whole agent is a setting for no real workflow.
"Ask everything" buries the user in approvals for trivial reads; "do everything" hands a single blank check to tools that should never auto-execute. Real workflows are mixed-risk, so a binary autonomy switch is wrong at both ends at once — too cautious for the safe actions, too reckless for the dangerous ones.
Autonomy Dial moves the grain of permission from the agent to the individual tool and action: auto-run this, ask-before-act on that, denylist the rest. Because each gate is labeled and durable across sessions, trust is calibrated by category once and the agent adapts on its next call rather than re-litigating it every run.
Trust an agent the way you trust a colleague — per task, not all at once.
DOPE Lens
How this pattern improves AI interaction quality.
Observability
Active permission scope visible per tool, agent, and action
Directability
Toggle, allowlist, or denylist any tool at any time — the agent adapts on the next call
Predictability
Each gated action is gated the same way every run — no surprise auto-execution
Explainability
Each permission carries a label naming the tool, the scope, and what triggers a gate
Fit
Where it earns its place.
Use it when
Delegation needs to vary by task type — research can be auto, sending mail must be approved.
Avoid when
The workflow is simple enough that one global mode is sufficient and the gate UX adds friction without insight.
In the wild
Where this pattern already ships.
Real implementations from products users already know. Each entry shows how the pattern got applied — what shipped, where it sits in the product, what it traded off.
Binary autonomy settings are too blunt for mixed-risk workflows where some tools should run free and others should never auto-execute.
Pattern
Let users grant per-tool / per-action autonomy — auto-run, ask-before-act, or denylist — instead of one binary toggle. The 2026 reality is gates more than continuous sliders.
Why Chat Breaks
Conversation alone cannot express durable, per-tool permission boundaries that survive across sessions.
Risks
Too many knobs can offload system design complexity onto the user; too few collapse back to the binary toggle this pattern is meant to replace.
Common questions
What people ask about this pattern.
What is an Autonomy Dial?
+
An Autonomy Dial is a per-task setting that lets users choose how much an AI may do on its own — from copilot (suggests only) to delegate (asks before acting) to autonomous (acts and reports back). It replaces a single global toggle with task-specific defaults users actually trust.
When should I avoid an Autonomy Dial?
+
Avoid the pattern when one global mode genuinely suffices for the workflow. Adding per-task knobs without a real reason offloads system design complexity onto the user and erodes trust.
How does this support directability?
+
Autonomy is no longer binary — users can keep manual control where stakes are high and delegate where they are not, all from a single visible surface that survives across sessions.
Grounded in Research
Not just opinion.
The papers, guidelines, and usability studies this pattern rests on.
Empirical study: users value agency preservation but accept assistance when it improves outcomes. Nuanced data on mixed-initiative perception — calibration evidence for any control-aware UX argument.
Six-chapter first-party guide. Its Explainability + Trust and Feedback + Control chapters map almost one-to-one onto our Trust & Verification and control patterns; its confidence-display guidance ("determine if, when, and how to show confidence") is the softer default our confidence-signals avoidWhen encodes.
Related Patterns
Curated connections, not just neighbors.
Hand-picked relationships — what this pattern pairs with in practice, what it replaces, what it builds on, and what builds on it.
