Licensing
AI Training Data Licensing
This archive offers pre-AI baseline imagery with verified human-captured provenance. Authentic, labeled, and timestamped data for teaching models how to render the Arctic and Alaska correctly.
Why This Data Is Valuable
- Pre-AI baseline — Archive dates to 2007. A physical-world record before synthetic content dominated the internet.
- Niche complexity — Aurora physics, midnight sun, spawning salmon, glacial dynamics. High-complexity environments where AI struggles.
- Timestamped provenance — EXIF, IPTC DigitalSourceType, C2PA signing. Machine-verifiable human capture.
- Geographic coverage — Geocoded locations, place metadata, structured collections.
Archive Overview
- Total images
- 8 items across 4 collections
- Date range
- 2016–2026
- C2PA signed
- 0 (growing with new imports)
- Categories
- kenai-peninsula, arctic-north-slope, fisheries-rivers, aurora-night-sky, communities-culture, glaciers-ice, establishments
Licensing Tiers
Per-image
Single images or small sets for targeted model improvement.
Collection
Full collections (e.g. aurora, glaciers, wildlife) with metadata.
Full archive
Complete archive access with images + metadata (JSON/JSONL).
Data formats: images (WebP/JPEG), metadata JSON, or JSONL for training pipelines.
Contact for AI Training LicensingMachine-Readable Feed
AI companies can programmatically query licensing availability at /api/licensing.json.