ShareLiveFraud is the public beta of a larger system for checking claims — every alert is AI-checked, independently battled, human-approved, and receipted.
Every alert is held to a standard — sourced, independently battled, human-approved, receipted. Someone keeps that standard. That's the Keeper.
What each of those words means
- AI-checked — the claim is read against the sources, and what cannot be supported is dropped.
- Independently battled — models from different companies review it separately, and they disagree in public on the receipt.
- Human-approved — a named person affirms six things about the evidence, and nothing publishes without all six.
- Receipted — every alert has a page showing its sources, its dissent, and what we could not confirm.