Human-in-the-loop review

Human-in-the-loop review is a way of working in which a qualified person checks, and can amend or reject, AI output before it is relied on or sent to anyone.

Last updated

What it means for a UK law firm

Human-in-the-loop review keeps professional judgement with the solicitor. In a law firm, human-in-the-loop review means AI output is treated as a draft until a named person has reviewed it, and that person is accountable for the result.

Human-in-the-loop review is only as good as the reviewer's checks. A review that reads an AI summary but not the sources it relies on will not catch an AI hallucination. Effective review looks at the cited passages, records who reviewed what, and stops work being exported or sent until review is complete.

Example

Illustrative scenario

A firm's rule is that no AI-generated report leaves the matter until a solicitor has reviewed every section. A trainee's report is held back because one section has been marked as needing attention, and is released once the point is resolved.

How Chancery handles it

In Chancery, reviewers mark each section of an output Reviewed or Needs attention. Approval stays locked until every section has been reviewed, approved outputs export to Word or PDF with a full citations appendix, and unapproved work stays in Chancery; see Chancery Verify.

Related terms

This glossary entry is general information, not legal advice. See the full legal AI glossary for other terms.

Give your firm the same advantage as the top 20, at a price that makes sense for yours.