DisMyth
Principles Process Corrections Funding Open the app
Methodology & Standards

How DisMyth reaches a verdict — and holds itself accountable.

This page is DisMyth's public standard of practice. It documents how we check claims, cite sources, correct our mistakes, and stay non-partisan — written to align with the IFCN Code of Principles and the EFCSN Code of Standards.

◇ IFCN-ALIGNED · APPLICATION IN PROGRESS LAST UPDATED · JUL 2026
The five commitments

Every IFCN principle, mapped to a feature you can see.

The IFCN Code rests on five commitments. For each, here is exactly how DisMyth delivers it in the product — not as a promise, but as a mechanism.

01
Non-partisanship & fairness
We apply the same verification standard to every claim, regardless of who made it or which side it favours. We do not concentrate checking on one party, and we take no policy positions on the issues we check.
IN THE APP
Per-claim bias & neutrality meter (Left ↔ Center ↔ Right) with a framing note on every verdict card.
02
Transparency of sources
Readers can verify our findings themselves. Every verdict lists its underlying evidence with links, so the chain from claim to conclusion is fully traceable.
IN THE APP
Weighted evidence panel — each source typed (Official / Scientific / Forensic / News) and credibility-scored, with an origin & spread trace.
03
Transparency of funding & organization
We publicly disclose who owns and funds DisMyth, and demonstrate that no funder — commercial, state or political — controls our editorial verdicts.
ON THIS PAGE
See the Funding & ownership section below for our revenue sources and editorial-independence firewall.
04
Transparency of methodology
We explain how we select, research and reach a verdict on claims, and how readers can submit claims for checking. No single model is a black box.
IN THE APP
Multi-AI consensus engine — Claude, GPT-4o, Grok & Gemini vote, and the agreement % is shown. See the Process below.
05
Open & honest corrections
When we get something wrong, we correct it clearly and promptly, and log the change so the record is auditable.
IN THE APP
Contested → human-review flow holds disputed claims from broadcast; see the Corrections policy.
How a verdict is made

From claim to conclusion — the exact pipeline.

01 · INTAKE
Selection & framing
A claim enters via user submission or automated monitoring. We restate it as a single, checkable factual assertion — stripping opinion and rhetoric.
02 · RETRIEVAL
Evidence gathering
We pull primary evidence — official records, scientific literature, forensic analysis, reputable reporting — and weight each source by type and reliability.
03 · CONSENSUS
Multi-model assessment
Four independent AI models assess the claim against the evidence in parallel. We record each model's call and compute an agreement score.
04 · ADJUDICATION
Verdict or escalation
High agreement → a labelled verdict with confidence. Disagreement → the claim is marked Contested and held from broadcast for human review.
VERDICT LABELS
FALSE TRUE MISLEADING UNVERIFIABLE CONTESTED · IN REVIEW
Every label carries a confidence percentage and the model agreement count. A verdict is never published on a single model's word alone.
Corrections policy

We publish our mistakes as loudly as our verdicts.

A commitment to open and honest corrections is non-negotiable under both the IFCN and EFCSN codes. This is ours.

How to flag an error
Anyone can dispute a verdict from its card, or email corrections@dismyth.app. Every dispute is logged and acknowledged.
What happens next
A disputed verdict is re-opened, re-checked against fresh evidence, and — if wrong — reversed. Broadcast is paused while it is Contested.
How corrections appear
A corrected verdict card is labelled Corrected, shows the old and new call, the date, and a plain-language reason. Nothing is silently edited.
Auditable record
Every correction is kept in a public log so the network's accuracy — and our willingness to fix errors — can be independently reviewed.
Funding & ownership

Who pays for DisMyth — and who doesn't control it.

IFCN signatory status cannot go to organizations whose editorial work is controlled by a state, party or politician. Here is our full disclosure.

REVENUE SOURCES
DisMyth is funded by DisMyth+ individual subscriptions and Pro plans for newsrooms & NGOs. Core claim-checking is free as a public good. No political party or government funds our operations.
EDITORIAL FIREWALL
No advertiser, subscriber or Pro customer can influence a verdict. Verdict logic is governed by the methodology above; the human-review board operates independently of the revenue team.
OWNERSHIP
DisMyth is an independent project created by Rehan Imam Waris. Legal registration as a dedicated fact-checking entity is underway as part of our IFCN application.
CONFLICTS OF INTEREST
Team members recuse themselves from any claim in which they have a personal, financial or political interest. Recusals are noted on the affected verdict.
Road to certification

Where we are on the path to becoming an IFCN signatory.

Signatory status is granted to legally registered fact-checking organizations that regularly publish non-partisan reports, after assessment against 31 criteria by an independent assessor. Our status:

DONE
Published methodology, corrections policy and funding disclosure (this page)
DONE
Non-partisan, source-transparent verdict system live in the app
IN PROGRESS
Build a public track record — publish fact-checks regularly for the required period
IN PROGRESS
Legal registration as a dedicated fact-checking organization
NEXT
Submit the IFCN application (US$350 assessment fee) for review against the 31 criteria

Standards you can hold us to.

Spot something we got wrong? Dispute it in the app — that's how the system is meant to work.

DisMyth
DISMISS THE LIE · CREATED BY REHAN IMAM WARIS
HomePrinciplesProcessCorrectionsOpen app