Skip to content

The Loop Model

Three ways humans govern AI

Loop separates three things often collapsed into one phrase: reviewing outputs, monitoring systems and owning outcomes.

Why Human-in-the-Loop became overloaded

Human-in-the-Loop made most sense when AI systems produced outputs for humans to use. The governance model was simple: a person reviewed before anything happened.

As systems gained autonomy, taking actions, routing work, creating records and triggering workflows, the same phrase was stretched to cover monitoring, process design, and accountability. Organisations could claim to be "in the loop" while meaning very different things.

Loop gives clearer language by naming three distinct patterns: Human-in-the-Loop (a human reviews each output before action), Human-on-the-Loop (humans monitor the system and intervene on exceptions), and Human-Accountable-for-the-Loop, HAL (a named person owns the system, its authority, its controls and its outcomes).

Each asks a different question and fits a different class of workflow.

The three models

Review

Human-in-the-Loop

A human reviews or approves individual AI outputs before action is taken.

AI → Human Review → Action

Human role

Reviewer

Key question

Did a human review the decision?

Example

An AI drafts a clause summary. A lawyer reviews it before relying on it.

Best for

  • +Drafting
  • +Summarisation
  • +Legal research
  • +Low-volume triage
  • +Internal recommendations
  • +Assistant-style tools

Strengths

  • +Clear human decision point
  • +Easy to explain
  • +Familiar governance pattern
  • +Strong fit for legal work where judgement remains central

Limitations

  • −Does not scale well to high-volume workflows
  • −Can become performative if review is rushed
  • −May create false comfort if the reviewer lacks context
  • −Slows down workflows where automation is the point
Review health checklist Workflow calculator
Monitor

Human-on-the-Loop

An AI system operates within defined boundaries while humans monitor behaviour, investigate exceptions and intervene when required.

AI → Action ↓ Human Monitoring

Human role

Monitor

Key question

Is a human monitoring the system?

Example

A regulatory monitoring system identifies potential changes and routes unusual or high-impact items for review.

Best for

  • +Alerting systems
  • +Risk monitoring
  • +Operational dashboards
  • +Compliance screening
  • +Exception-based workflows
  • +Medium-volume automation

Strengths

  • +Better suited to scale than Human-in-the-Loop
  • +Allows automation while preserving oversight
  • +Supports exception-based intervention
  • +Useful where normal operations are predictable

Limitations

  • −Requires strong monitoring
  • −Humans may miss weak signals
  • −Intervention thresholds must be well designed
  • −Accountability can become unclear if ownership is not defined
Monitoring design checklist Workflow calculator
Own

Human-Accountable-for-the-Loop(HAL)

A human owns the design, authority, controls, evidence and outcomes of an AI-enabled workflow, even where the AI system acts without individual human review.

Owner → Policy → AI Workflow → Action → Evidence → Review

Human role

Owner

Key question

Who owns the system that made the decision?

Example

A matter triage agent classifies incoming legal requests, routes work, creates records, escalates risk and logs evidence. No human reviews every decision, but a named owner is accountable for the workflow, its controls and its outcomes.

Best for

  • +Agentic systems
  • +Multi-agent workflows
  • +Workflow orchestration
  • +Large-scale triage
  • +Obligation tracking
  • +Autonomous operational systems
  • +High-volume decision workflows

Strengths

  • +Designed for scale
  • +Focuses on real accountability
  • +Supports agentic workflows
  • +Forces clarity on ownership, authority, limits, evidence and escalation

Limitations

  • −Requires mature governance
  • −Not suitable where accountability is vague
  • −Needs technical logging and review processes
  • −Higher setup burden than basic review workflows
Explore the HAL framework Take the HAL assessment

From review to monitoring to accountability

The three models are not a maturity ladder. They are different governance patterns for different workflow types. Human-in-the-Loop remains the right choice for assistant-style AI where a human must review each output.

As volume and autonomy increase, review alone stops scaling. Human-on-the-Loop fits workflows where the system operates within boundaries and humans intervene on exceptions.

When the system takes action at scale, routing work, creating records and triggering workflows, individual review is no longer practical. Human-Accountable-for-the-Loop shifts governance from the decision to the system: who owns it, what are its limits, and how is it evidenced and reviewed.

How autonomy changes the governance requirement

A drafting assistant and an autonomous triage agent are not held to the same standard. The more a system can do without human review, the more explicit ownership, authority, limits, evidence and escalation must be.

Loop helps you identify which pattern applies. HAL helps you assess whether a workflow has the accountability maturity to operate at its level of autonomy.

Why HAL requires a deeper framework

Naming an owner is necessary but not sufficient. Agentic workflows need explicit authority, enforced limits, escalation paths, immutable evidence, ongoing monitoring, periodic review, and clear liability allocation.

HAL provides eight domains, a scoring system, and practical templates for workflows where individual review does not scale. It addresses a different class of problem from Human-in-the-Loop.

HAL is not a label for autonomy. It is a test of whether autonomy has been properly bounded, evidenced and owned.

Read the HAL framework →

When each model fits

Workflow type Recommended model
Drafting a memo Human-in-the-Loop
Summarising a contract Human-in-the-Loop
Monitoring regulatory updates Human-on-the-Loop
Screening incoming matters Human-on-the-Loop or HAL
Drafting a client communication Human-in-the-Loop
Routing client requests HAL
Updating matter records HAL
Triggering external communications HAL with approval gates or strict controls
Filing regulatory documents HAL only with high score and approval gates

Practical decision tree

  1. Question 1

    Does the AI system only generate outputs for human use?

    If yes: Human-in-the-Loop likely fits.

  2. Question 2

    Does the AI system operate continuously or at scale, but humans intervene when needed?

    If yes: Human-on-the-Loop likely fits.

  3. Question 3

    Does the AI system take action, trigger workflows, create records, route work or affect outcomes without review of every action?

    If yes: Human-Accountable-for-the-Loop likely fits.

  4. Question 4

    Could failure create legal, financial, regulatory, reputational or client impact?

    If yes: Assess against HAL before deployment.

Use the workflow calculator See worked examples