ET Workbench

Employment Tribunal decisions, read by AI

Building AI tools from Employment Tribunal decisions

Practical questions about Employment Tribunal claims, answered from more than 134,000 published judgments using AI that runs on a single desktop computer. This site explains how each tool works, what it finds, and where its limits are.

Read the articles →Search the injury to feelings awards →

Line chart: judgments in the corpus by the year the claim was presented, 2010 to 2025. 85 in 2010, a peak of 23,286 in 2019, then 5,407 in 2025. A dashed line marks the start of the online register in 2017.
Judgments in the collection by claim year. Recent years are lower partly because many of those claims have not yet reached a judgment.
  • 134,988judgments in the corpus
  • 2,527injury to feelings awards
  • 12,857disability discrimination judgments
Where 2,527 injury to feelings awards sit within the Vento bandsEach dot is one award, placed by reference to the bands for the year the claim was presented. 7 awards without a readable claim year are not shown. Open the viewer to search them.
Strip chart of 2,527 injury to feelings awards by position within the Vento bands: 1,236 in the lower band, 1,191 in the middle band, 86 in the upper band and 7 above it. 7 awards without a readable claim year are not shown.
Lower band
1,236
Middle band
1,191
Upper band
86
Above the upper band
7

Above upper: more than the upper band limit for the year the claim was presented.

Latest

Trends in the tribunal · 4

What tribunals see when parties use AI

Parties' use of AI now appears regularly in Employment Tribunal judgments. Tribunals accept that people use it, but hold them responsible for what it produces.

11 min read

  • Trends in the tribunal · 3

    Out of time: how often tribunals let late claims in

    What the published Employment Tribunal judgments show about extending time under the 'reasonably practicable' and 'just and equitable' tests: how often late claims get in, what made the difference, and where representation made a difference.

    11 min read

  • Trends in the tribunal · 2

    Disability in the tribunal: a viewer for 12,857 judgments

    Every disability discrimination judgment in the published register, the impairments relied on, and what the tribunal decided about each, searchable by condition, outcome and the part of the test that failed.

    4 min read

  • Trends in the tribunal · 1

    ADHD and autism in the tribunal

    How often ADHD and autism are relied on as disabilities in Employment Tribunal claims, how often respondents concede them, and how often tribunals find they meet the statutory test.

    6 min read

  • Valuing injury to feelings · 1

    Injury to feelings: building the awards database

    How injury to feelings awards were extracted from every published Employment Tribunal judgment, using a pattern-matching pass, a language model running locally, and checks of every figure against the text of the judgment.

    10 min read

All posts →

The series

  1. Foundations

    1. 1AI in the Employment Tribunal
    2. 2Building the corpus: 134,000 Employment Tribunal judgments
    3. 3Data protection and the judgments

  2. Valuing injury to feelings

    1. 1Injury to feelings: building the awards database
    2. 2Injury to feelings: a viewer for more than 2,500 published awards

  3. Assessing a claim

    1. 1Local and cloud AI: what the difference means for employment advice

  4. Trends in the tribunal

    1. 1ADHD and autism in the tribunal
    2. 2Disability in the tribunal: a viewer for 12,857 judgments
    3. 3Out of time: how often tribunals let late claims in
    4. 4What tribunals see when parties use AI
  5. Coming later

    Testing AI on employment law

    A blind-marked comparison of how general-purpose AI services and a local tool handle realistic employment law problems.

About the series →