Trust Centre
Parliament People Trust Centre
A public guide to how Parliament People sources parliamentary information, uses AI and editorial review, handles corrections, and records material transparency updates.
Public Policy Docs
Trust And Safety Mission
The operating mission behind Parliament People's public trust work.
How We Use AI
Where AI appears, what it is for, and what it is not.
How Provenance Works
How source records, generation records, and AI labels fit together.
Prompt Orchestra Transparency
How Prompt Orchestra supports versioned, inspectable generation.
Corrections, Redress, And Appeals
How users can challenge content and how reports, corrections, and reconsideration are reviewed.
Editorial Integrity Code
How Parliament People handles AI-assisted civic explanation and editorial judgement.
Community Guidelines
Behaviour expectations for current reporting flows and future interactive features.
Known Limitations
What Parliament People can and cannot claim about generated summaries.
Public Operating Docs
Operations Index
Inventories, risk taxonomy, review queues, claims, publication, disclosure, and correction gates.
Content And Surface Inventory
The public map of content classes, surfaces, sources, and known trust gaps.
Risk Taxonomy
Risk families, severity model, and routing assumptions for trust and safety review.
Review Queues And Lines Of Defence
How reports, corrections, source issues, and escalations should be routed.
Product Commitments
- AI-assisted content should be source-grounded and attributable to parliamentary records.
- Generation records should be reviewable where a public Prompt Orchestra run record is available.
- Users should be able to report errors, missing context, unclear labels, and other concerns.
- Material corrections, policy updates, source issues, and system changes should be considered for public transparency logging.
- The official parliamentary source remains authoritative.