For publishers, journalists, creators, researchers, and students

Vefy Reporting Standard™

Don’t ask people to trust you. Show your work. The Vefy Reporting Standard™ is a public framework for reporting that is easier to inspect, harder to misread, and more defensible when the evidence is challenged.

Primary-source first Attribution stays explicit Evidence trail stays reproducible Trust Index™ follows evidence, not the other way around

Open the Vefy Evidence Checklist™ See before and after examples

Phase 1

A dedicated standard that fits Vefy without distracting from the consumer report flow.

What Vefy already does today

Investigates public narratives after publication with Verdict, Trust Index™, Argument Engine, receipts, and visible revisions.

What this adds

A pre-publication evidence framework for making stories more inspectable before they spread.

What it does not do

Guarantee a future score, sell favorable treatment, or replace Vefy’s independent investigation.

Cultural idea

Move from “trust me” to “check me.”

If a story relies on a filing, dataset, transcript, study, complaint, or statement, the reader should be able to inspect the record directly whenever practical.

Product fit

Teach stronger reporting without turning Vefy into a publisher perk.

The standard can help a small newsroom or independent journalist show strong evidence practices without implying that Vefy endorses the conclusion.

Independence

Evidence provided is not the same thing as Vefy verified.

Publishers can show the records they relied upon. Vefy still has to test whether those records actually support the claim being made.

Core principles

Make the evidence as easy to inspect as the claim.

The standard is built around evidence transparency, claim-source discipline, and honest uncertainty. It should make stronger reporting feel concrete, not abstract.

01

Show the primary source

Prefer the filing, complaint, dataset, transcript, study, official statement, or original video over another article describing it.

02

Attribute consequential claims

Readers should know who is making the assertion, whether it is a plaintiff, regulator, company, witness, court, reporter, or analyst.

03

Separate allegation from established fact

An investigation is not a charge. A charge is not a conviction. A lawsuit is not a judgment. Reporting should preserve those distinctions.

04

Contextualize numbers

Important numbers should carry the source, timeframe, denominator, comparison period, and material caveats that keep them from misleading by omission.

05

Preserve quote context

Whenever possible, link the transcript, interview, full video, hearing, or surrounding statement so a quote can be inspected in context.

06

Disclose material uncertainty

It should be normal to say when something is developing, not independently confirmed, disputed, preliminary, or not fully established by the available record.

07

Make the trail reproducible

Another person should be able to follow the path from claim to source to record to context without guessing how the story got there.

08

Correct in public

Material corrections should not silently disappear. Good reporting leaves a visible record of what changed, when, and why.

Visual evidence chain

The shorter and clearer the path, the easier the story is to inspect.

ClaimWhat is being said?

SourceWho is saying it?

Primary recordWhere is the underlying document?

CorroborationWhat else supports or challenges it?

ContextWhat changes the interpretation?

ConclusionWhat does the evidence support right now?

Before and after

Better evidence practice should be obvious in seconds.

The goal is not more links for the sake of links. The goal is clearer attribution, more inspectable support, and fewer claims that outrun the record.

Example 1

Allegation vs. finding

Before

Regulators found the company committed fraud.

Evidence-ready

Regulators opened an investigation into fraud allegations involving the company.

Problem: the linked record announces an investigation, not a formal finding.

Example 2

Numbers need context

Before

Complaints increased 200%.

Evidence-ready

Complaints increased from 3 to 9 during the reporting period, according to the agency’s public dataset.

Context changes whether a number sounds explosive, limited, or simply early.

Example 3

Linking a source is not enough

Before

According to an SEC filing, the company closed the acquisition.

Evidence-ready

The company said in its SEC filing that it expects the acquisition to close next quarter. The filing does not state that the deal has already closed.

A highly authoritative source can still be overstated, misread, or quoted beyond what it actually says.

Vefy Evidence Checklist™

Before you publish, pressure-test the evidence trail.

☐Are consequential factual claims clearly attributed?

☐Are primary records linked where available?

☐Do the linked records actually support the characterization?

☐Are important numbers sourced and contextualized?

☐Are quotes connected to original context where practical?

☐Are allegations identified as allegations?

☐Are reporting, analysis, inference, and opinion distinguishable?

☐Are material uncertainties disclosed?

☐Is breaking or original reporting labeled appropriately?

☐Could another person reproduce the evidence trail?

Important distinction

Source authority and claim support are different tests.

Source authority

How authoritative is the source itself? Filing, court record, dataset, study, official statement, transcript.

Claim support

Does that source actually support the specific statement being made, or has the source been stretched beyond what it says?

Guardrail

Links alone do not equal credibility.

Relevant source linked ✓

A primary source is visible and inspectable.

Claim still unsupported ✕

The source may be mischaracterized, incomplete, contradictory, or too weak for the conclusion drawn from it.

No source-count shortcut

Twenty weak citations do not automatically beat one decisive primary record, and repeated commentary does not equal independent corroboration.

Breaking and original reporting

Early reporting should be evaluated honestly, not punished automatically.

Sometimes the first good report appears before corroboration exists. Vefy should distinguish original reporting, developing stories, emerging corroboration, and mature records without pretending that early evidence is the same as settled evidence.

State 1

Original reporting

The publication may be first. Underlying evidence still matters more than whether others have echoed it yet.

State 2

Developing story

Useful records exist, but uncertainty remains visible and important pieces may still be missing.

State 3

Corroboration emerging

Independent confirmation starts to arrive, but may still be uneven, partial, or derivative.

State 4

Well corroborated

Multiple independent records and durable primary evidence support the reporting with fewer unresolved gaps.

Corrections and transparency

Good corrections stay visible.

CORRECTION — AUGUST 8, 2026

Earlier version: $14.2 million

Corrected: $12.4 million

Reason: underlying filing was updated.

Upcoming direction

Vefy PreCheck™ is a strategic direction, not a public promise.

Possible future role

Help a newsroom, student paper, company, or researcher identify what would make a draft more evidence-ready before publication.

What it would evaluate

Attribution, primary sourcing, claim-source alignment, numerical context, uncertainty disclosure, and reproducibility.

What it would not do

Guarantee a future Trust Index™, promise a favorable Vefy Verdict, or create a pay-to-trust system.

Final principle

The score follows the evidence. The evidence does not follow the score.

Vefy should never tell publishers how to game a number. It should help them make reporting more transparent, more inspectable, and more faithful to the underlying record. If that leads to stronger public assessments later, it should happen because the reporting is actually better supported.

Read Methodology Open Trust Center Explore Evidence Submission