This document is an implementation specification for an Auto-Registration Permanent Archive Service — a system that automatically preserves web content at the moment of publication, with cryptographic tamper-evidence and legally admissible timestamping. Key technical components: WARC-format full-page capture (ISO 28500) via Headless Chrome, RFC 3161 trusted timestamping (multi-TSA, parallel), OpenTimestamps anchoring to Bitcoin blockchain, independent observation nodes across 3 geographic regions for cross-validation, platform integration via OAuth 2.0 and Webhooks (WordPress, note, Wix, Medium, Ghost), PDF + JSON tamper-proof certificates with public verification endpoint, and genre-specific persistent identifiers (vio.paper, vio.text, vio.art, etc.). This specification addresses a critical gap in the AI era: the inability to prove "this content existed in this exact form on this date." It serves AI provenance verification, academic citation integrity, and legal evidence use cases. The service is owner-initiated, not third-party archiving — fundamentally different from Internet Archive or existing cache services. This work also positions itself in relation to C2PA (Coalition for Content Provenance and Authenticity), identifying areas where WARC-based capture-time authenticity complements C2PA's current coverage.
Just published on Zenodo 🗂️
Auto-Registration Permanent Archive Service — proves "this content existed in this exact form on this date."
Owner-initiated. Tamper-proof. Blockchain-anchored.
For the AI provenance gap.
zenodo.org/records/1957...
#WebArchiving #AIProvenance #C2PA #OpenTimestamps #WARC
