Honest, plain-English description of how ZARN processes your content during the pilot.
ZARN uses AI-assisted analysis to assess submitted content for protective security risks. Output is decision-support, not a formal threat assessment. Every finding requires human review before action.
Uploaded images and documents are deleted automatically 24 hours after upload. The assessed text, extracted metadata and resulting findings are retained as part of the scan record, together with the file name, type and size, so that reports, evidence packs and returning-finding detection remain accurate. Webpage and free-text scans involve no uploaded file.
Scan content may be processed by configured AI infrastructure and subprocessors engaged by ZARN to deliver the service. A current list of subprocessors is available from your ZARN pilot point of contact.
ZARN maintains internal technical audit logs — including AI provider and model metadata — for security, quality assurance and troubleshooting. Customer-facing reports do not expose these low-level technical details.
Internal, private and unsafe network addresses are blocked. You may only scan public URLs you are authorised to assess.
If a ZARN super-admin needs to view a customer scan (for support, abuse review or quality investigation) they must provide a written reason. The reason and access event are written to an immutable access log before the content is shown.
Generated reports are AI-assisted. Review and validate before distributing internally or externally.
ZARN LTD is a private limited company registered in England and Wales under company number 16431762. Registered office: 71-75 Shelton Street, Covent Garden, London, United Kingdom, WC2H 9JQ.