Source & Verification Policy

For every source we consider its authority, origin, independence and how close it is to the facts. The three broad families are shown in the table at the top of this page.

A source’s category doesn’t settle how reliable it is. Each one is judged in context.

When reporting is about a lawsuit, regulatory action, company filing or government decision, we normally look for the underlying record rather than rely only on someone else’s description of it.

If a news report says a lawsuit was filed, we try to read the court record. If a regulator is said to have made a finding, we look for the actual decision, order or notice.

That way we don’t repeat another publisher’s mistake or spin.

Different court documents prove different things.

An allegation in an official filing doesn’t become a decided fact just because it sits on a court docket.

We keep these apart:

If a settlement says a party neither admits nor denies the allegations, we don’t describe it as an admission unless something else reliable supports that.

Company registries and statutory filings can help establish:

What a filing legally means depends on the jurisdiction and the type of filing.

• A registered address doesn’t necessarily show where someone lives.
• Being a director doesn’t automatically make someone responsible for everything a company does.
• Owning shares doesn’t necessarily mean running the business.

Every material piece of evidence goes through the eight checks set out near the top of this page — authenticity, provenance, authority, date, context, completeness, current status and relevance.

Looking professional or being widely shared doesn’t count as verification.

When a serious allegation comes from one source, we look for independent support wherever we reasonably can.

Ten websites copying one original article are not ten independent sources. Real corroboration can come from:

The more serious the allegation, the more checking it needs before we publish.

Documents can come from subjects, lawyers, whistleblowers, staff and former staff, investors, the public and others. Before relying on one, we may ask:

• Who sent it
• How they got it
• Whether an original exists
• Whether its metadata or origin can be confirmed
• Whether it has been altered
• Whether signatures or identifying marks can be authenticated
• Whether independent records back up what it says

Being a member doesn’t give your documents extra weight.

Submit evidence for review

Screenshots, messages, emails, social posts, websites and other digital material get careful checks before we rely on them. Where it matters, we may look for:

A cropped screenshot with no context is not normally enough to prove a serious allegation.

Audio, video and photos may need checking for:

Generative AI and synthetic media make authentication more important than ever.

If we can’t reasonably establish that something is genuine, we either say so or don’t rely on it for a material conclusion.

We may use technology to help with search, organising documents, translation, matching names to entities, comparison and other research tasks.

An answer, summary or inference from an AI system is not evidence just because a system produced it.

Any material claim found with automated tools is traced back to reliable sources before publication.

We never knowingly present AI-generated quotes, documents, photos, sources or evidence as genuine.

We may use lawful open-source methods to find and analyse public or lawfully available information, including:

Anything found this way still has to be verified. A connection suggested by a software tool is not automatically real.

Where appropriate, investigators may lawfully visit places to check addresses, businesses, operations or other things relevant to an investigation. They keep three things apart:

We never authorise trespass, unlawful surveillance, bribery, theft or any other unlawful way of getting evidence.

With information from interviews, researchers consider:

• Whether the person knows it first-hand
• Their relationship to the subject
• Whether the account is from the time or recalled later
• Any financial, professional or personal conflicts
• Whether documents support what they say
• Whether other witnesses can confirm the important points

There can be good reasons to protect a source’s identity.

When we rely on a confidential source, the people responsible should, where practical, know who the source is, how they know what they say, what might motivate them, and their link to the matter.

Material allegations from confidential sources are independently corroborated wherever reasonably possible.

Source protection depends on the law. We can’t promise it is absolute in every jurisdiction or situation.

A source might be:

A conflict doesn’t make information false. It does mean the claims may need extra scrutiny and independent support.

We never let a source’s personal dispute stand in for evidence.

Documents or testimony that contradict an allegation get the same serious consideration as material that supports it.

Reliable evidence of mistaken identity, dismissal, acquittal, reversal, factual error or another clearing development must be weighed when deciding what the record can fairly say.

We don’t verify only the evidence that fits an adverse story.

Material USA Scammers records can carry a numbered source list. Where appropriate, each entry can show:

That lets readers trace each material claim back to what supports it.

Where appropriate and lawful, we may keep the source material behind significant reporting so we can:

• back up what we published;
• review later challenges;
• keep a version history;
• assess corrections;
• deal with legitimate legal issues; and
• understand later developments.

We keep it in line with privacy, data-protection, security, confidentiality and other legal duties.

We may receive material that is confidential, restricted, privileged, covered by a court order or otherwise legally sensitive. It gets proper legal and editorial review before any use or publication.

We don’t assume public interest automatically beats confidentiality, privilege, privacy, reporting restrictions or other legal duties.

If material looks altered, forged, selectively edited or otherwise unreliable, publication is put on hold until the concern is resolved.

Knowingly sending us fabricated evidence can lead to the submission being rejected, access being restricted or other action.

Where warranted, we may take legal advice on next steps.

Where reasonable, we may ask a subject to confirm or comment on documents about them before we publish.

A subject’s denial doesn’t automatically cancel out independently verified evidence.

And a document isn’t assumed genuine just because the subject didn’t reply.

Agencies, courts, registries and regulators sometimes correct or amend their records.

If credible evidence suggests an official source is wrong, we investigate rather than repeat it mechanically.

If an official source is disputed, we describe the dispute accurately where it matters.

Records about the same person or company can differ from country to country. Names get transliterated differently, company identifiers vary, courts use different terms and access to official information differs widely.

So we confirm cross-border links with identifiers and context — never with name similarity alone.

We don’t treat “Asia,” Europe or any other region as one legal or evidence system.