Made by AI, held to account
Lumidawn is operated end-to-end by AI. That is a promise about transparency, not a loophole around it.
Every piece is disclosed
Each video carries a persistent AI-generated label and, where supported, a C2PA content credential you can verify.
The newsroom is verified
News passes an independent verification gate — claims are matched against tiered sources and graded before anything is published. Editorial policy is protected and can only be tightened.
Ads never touch the news
Advertising is entertainment-only. A hard firewall keeps every paid placement out of and away from news.
How a video is made
- 01
Scout
Agents watch news feeds and market signals, clustering the stories worth covering.
- 02
Verify
Every claim is matched against tiered sources and graded. What fails the gate never proceeds.
- 03
Script
A script is written in the right voice, hedged to match the verification grade.
- 04
Produce
Voice, anchor and visuals are generated — cheapest assets first, inside the daily budget cap.
- 05
QC
A reviewer checks every claim against the fact sheet and catches audio-visual glitches.
- 06
Comply & publish
AI labels, C2PA credentials and a rights log are attached before publishing to every platform.
Editorial, safety and compliance policies are protected — the self-improvement loop may only tighten them, never weaken them.