Human-AI Collaboration

Escalation Pathways

Design structured escalation triggers and handoff mechanisms so agents can pause and ask for human guidance when they encounter ambiguity, conflicts, or decisions beyond their authorization - without breaking workflow or losing context.

What is Escalation Pathways?

Agents will encounter situations they can't handle - ambiguous instructions, conflicting information, high-stakes decisions they're not authorized to make, or tasks that exceed their capabilities. The agent needs a structured way to escalate to the human without breaking the workflow, losing context, or creating anxiety. This is different from simple error recovery because the agent hasn't failed - it's recognized its own limitations. The pattern defines four escalation types: confidence-based (uncertainty threshold), permission-based (authorization limits), conflict-based (contradictory information), and capability-based (task exceeds abilities). Each escalation preserves full context, includes a recommended action with confidence level, and allows the agent to continue from where it paused after the user responds.

Example: Notion AI - Scope Confirmation Before Bulk Action

Notion AI showing scope options for AI autofill with 'Try on this view' vs 'Fill all pages' choices

When asked to auto-fill a column with AI, Notion presents scope options — apply to this view only or fill all pages in the database. Gives users control over the blast radius before the AI acts across multiple items.

AI Design Prompt

Want to learn more about this pattern?

Explore the full pattern with real-world examples, implementation guidelines, and code samples.

View Full Pattern

Related Prompts from Human-AI Collaboration

Contextual Assistance

Human-AI Collaboration

Design an AI assistance interface with contextual suggestions, similar to Gmail Smart Compose or GitHub Copilot. Show subtle, inline suggestion chips that appear as the user types, positioned near the cursor or input area. Include: a text input area or editor, ghost text preview for AI suggestions (low opacity), inline suggestion chips with accept/dismiss actions, and keyboard shortcut hints (Tab to accept, Esc to dismiss). Style: Clean, modern, minimal interruption to user flow. Use subtle shadows, smooth transitions, and non-intrusive colors. Platform: Web application, responsive layout.

View Full

Human-in-the-Loop

Human-AI Collaboration

Design an interface where humans can review and approve AI suggestions before they're applied: Create a review card showing: 1. **AI Suggestion**: Display the AI-generated content/action with a confidence indicator 2. **Action Buttons**: Clear Approve/Reject/Modify options 3. **Context**: Brief explanation of why AI made this suggestion 4. **Override Option**: Allow users to edit or provide their own input Show visual distinction between AI suggestions (blue/purple) and human-approved items (green).

View Full

Augmented Creation

Human-AI Collaboration

Design an AI-powered creation interface where AI assists without taking over: Create a content editor showing: 1. **Creation Canvas**: Main workspace for user content (text, design, code) 2. **AI Suggestions Panel**: Side panel with AI-generated alternatives or improvements 3. **Accept/Modify Controls**: Easy buttons to accept, edit, or dismiss AI suggestions 4. **Collaboration Indicator**: Visual distinction between human-created and AI-suggested content 5. **Inspiration Mode**: Toggle for AI to generate multiple creative options Show the workflow: User creates → AI suggests improvements → User chooses → Final output. Include clear attribution showing what's AI-assisted vs. human-created.

View Full
Previous PromptAction Audit TrailView All PromptsNext PromptTrust Calibration