---
title: 'Justice Connect: Free-text legal-problem classification'
slug: justice-connect-legal-problem-classifier
stable_id: 88bf4a45891fbb14
company: Justice Connect
function_code: customer_service
pattern_codes:
  - intake_redesign
  - human_quality_loop
evidence_strength: reported
publication_tier: showcase
freshness: current
reviewed_at: '2026-08-23'
updated_at: '2026-08-23'
source_quality_summary: 1 independent, 1 primary; publication outcomes are reported.
caveat_summary: >-
  JusticeBench reports the operator's staged comparison. The retained non-AI path
  strengthens the comparison, but no independent statistical evaluation was published.
collections:
  - human-still-decides
  - queue-eliminated
bundle_version: 1.0.0
bundle_fingerprint: sha256:c23c6cc2b88153d008ea8fda928f632ce0011fc2d4c5036672a16e5d895bab93
canonical_url: https://brianletort.ai/transformations/justice-connect-legal-problem-classifier
---

# Justice Connect: Free-text legal-problem classification

People seeking legal help should describe their problem, not diagnose its legal category before they can ask for help.

Function: Customer service. Patterns: Intake redesign; Human quality loop. Evidence: reported.

Freshness: current. Reviewed: 2026-08-23. Updated: 2026-08-23.


Source quality: 1 independent, 1 primary; publication outcomes are reported.

## Before

1. **Help-seeker** — Selects a legal category from a fixed list, often choosing the wrong area or 'Something else'. (control: The applicant must understand the legal nature of the problem.)
2. **Intake staff** — Manually reclassifies uncategorized and misclassified applications. (control: Human triage and service eligibility rules.)
3. **Referral partner** — Reviews the referral and redirects matters that do not match its expertise or mandate. (control: Partner-specific service criteria.)

## After

1. **Help-seeker** — Describes the legal problem in ordinary language without selecting an area of law. (control: Disclosure, opt-out, and a retained non-AI path.)
2. **Legal-problem classifier** — Maps the free-text description to likely legal categories and confidence scores for routing. (control: The model classifies and routes; it does not provide legal advice.)
3. **Intake staff** — Reviews edge cases and works from cleaner, pre-categorized applications. (control: Human review remains available for uncertain or consequential cases.)
4. **Pro bono legal experts** — Compare model classifications with expert judgments and feed disagreements into refinement and retraining. (control: Standing expert-review loop.)

## Decision rights

The model proposes intake categories and routing. People may use the non-AI path, and humans retain review of edge cases and all legal-service decisions.

## Exception path

Users can opt out to a non-AI route. Low-confidence or consequential edge cases receive human intake review, while expert disagreement is recorded for model refinement.

## Outcomes

- **'Something else' selections during intake** (reported): 23% among users on the retained non-AI path → 9% among users offered the classifier. JusticeBench reports the operator's staged comparison. The retained non-AI path strengthens the comparison, but no independent statistical evaluation was published.

## Executive lesson

Move classification burden off the requester and into the intake system; keep experts on exceptions and model governance.

## Anti-pattern

Automating intake while forcing every user onto the new path removes the comparison group and hides whether routing actually improved.

## Questions for leaders

- Are requesters being asked to classify a problem they do not yet understand?
- Which cases require a human route regardless of model confidence?
- How does expert disagreement reach retraining and control changes?

## Sources

- [Justice Connect Legal-Problem Classifier](https://www.justicebench.org/project/justice-connect) — JusticeBench
- [Consensus: How pro bono lawyers help us maintain human oversight over our AI tools](https://justiceconnect.org.au/latest/blogs/2025/consensus-how-pro-bono-lawyers-help-us-maintain-human-oversight-over-our-ai-tools/) — Justice Connect