Neutrality
Appearing as a respondent in an employment tribunal case is a neutral fact. We do not label employers as "good" or "bad", we do not editorialise the merits of any particular case, and our headline figures (case counts, total awards) are descriptive, not comparative. A high case count can reflect employer size, sector, claim culture, or genuine workplace problems — a single number tells you none of those things.
Naming claimants
Claimant names are shown as written in the published judgment. Where the tribunal has anonymised parties (typically initials, or patterns like "A v B") we preserve that anonymisation. We do not attempt to re-identify anonymised claimants from external sources.
Naming respondents
Respondents are shown using the canonical legal name where we can establish one, falling back to the name written in the judgment. Where a respondent has multiple legal names across cases (e.g. re-incorporation, name change), we link them to a single employer page.
Awards
Award totals are taken from the remedy section of the judgment. Many UK tribunal cases are liability-only with the remedy decided at a separate hearing — those cases will appear with a recorded case count but no award until the remedy decision is also indexed. Settled, withdrawn, and struck-out cases never carry an award figure and are labelled accordingly.
What we don't publish
- Internal HR documents, witness statements, or correspondence referenced in a judgment but not part of its public text.
- Third-party identifying information that the tribunal redacted.
- Speculation about why a case was brought, settled, or appealed. Where motivation matters, the judgment text is the source.
Generative-AI use in editorial output
We use LLMs to help structure per-claim outcomes and key findings from judgment text. Every LLM-derived field is held to the underlying judgment as the source of truth — see Methodology. LLM-derived claims that cannot be cited back to the source PDF are dropped.