Trust and data handling

Built around traceability, cautious AI use and data minimisation.

Tribuno handles tribunal research as a source-linked professional workflow. The product is designed to show where information came from and what users still need to check.

Tribuno provides Employment Tribunal research assistance and legal information only. It does not provide legal advice, representation, or guaranteed outcome prediction.

Trust controls

What Tribuno makes visible

The immediate trust goal is practical: users should understand source provenance, AI involvement, limitations, and how live-case data is treated.

Source traceability

Tribuno is organised around source-linked records, citations, and provenance rather than free-form answers.

  • Selected ET decisions are stored separately from Lens findings.
  • Promoted Lens context uses `Lx` labels while ET decisions use `Sx` labels.
  • Memo detail keeps source lists and caveats visible.

Research-only boundary

The product supports professional research workflows and does not provide legal advice or prediction.

  • No win/loss prediction wording.
  • No automated solicitor positioning.
  • Memo outputs require independent professional judgement.

Data minimisation

Live-case content may be sensitive, so users should enter only what is needed for the research workflow.

  • Case Profiler is for research context, not unnecessary personal detail.
  • Account data controls explain export, deletion and support routes.
  • Support requests should use titles and routes rather than unnecessary sensitive detail.

AI with guardrails

AI-assisted features are used after retrieval and remain tied to selected sources and visible caveats.

  • Traceability details help explain how output was produced.
  • AI output should be treated as a review draft.
  • Users must check source decisions before relying on output.

Traceability example

Trust is shown through labels, provenance and checks

The product should make it clear whether a point came from a selected tribunal decision, promoted Lens context, or generated memo prose.

Source trail

Example traceability chain

This illustrative trail shows the trust pattern. Real workspace output should preserve the same separation while linking to actual selected sources and promoted context.

RefSource familyResearch roleTrace shownUser check
S1Selected ET decisionMemo citation sourceSource record, decision title, source document, selected passage and research-file membership.Open the decision and inspect the cited passage before using the memo wording.
L1Promoted Lens findingContext citationLens family, source passage, promotion action and label separate from tribunal comparables.Confirm the passage supports context only and does not replace professional legal research.
M1Generated memo sectionAI-assisted draftSelected-source labels, caveat, provenance metadata and visible source list.Treat the draft as research assistance and revise against the cited sources.

AI use

AI is used inside source-cited workflows

Tribuno should use AI for extraction, context organisation and memo drafting after retrieval, not as a free-standing employment-law advice tool.

After retrieval

Query-time drafting should use selected decisions, structured case context and source-linked Lens findings rather than rediscovering everything from raw text.

With metadata

Tribuno keeps enough traceability information to explain how research-support output was produced.

With review

AI-assisted output can be incomplete or wrong. Users should check source decisions and apply professional judgement.

Current access model

Account-based workspace access

Workspace access is account-based and available by request. Account activation, roles and access changes are handled by Tribuno support so research files and data controls stay in the right workspace.

Support model

Support is asynchronous and email-based. Tribuno does not promise phone support, live chat, demos, or consultancy-style onboarding for v1.

Live-case data

Start with sample or synthetic matters. If you use real facts, keep them focused on the research question and avoid unnecessary personal detail.

Review the product limits before requesting access.

The FAQ explains sources, AI use, data handling, and what Tribuno must not be used for.