Editorial policy

Purpose and scope

This policy applies to everything we publish: reporting, explainers, Warning List entries, firm files, timeline items and live desk updates. It binds staff, freelancers and anyone with editorial access to our systems. Where a rule below is enforced technically — in the database rather than by good intentions — we say so.

Sourcing

We prefer primary sources: regulator notices, court and tribunal records, company filings, official registers and documents readers send us. A single anonymous claim is not a story. A critical claim about a named party needs either a document, a public authoritative source, or two independent people in a position to know.

Every citation stored against a story carries the publisher, the URL and the date we last checked it. Where a source later changes its own page, the story records that the citation moved rather than quietly following it.

Accuracy and verification

We match firms on reference numbers and registered numbers, not on names, because clone operations exist precisely to make name matching fail. We distinguish, in the words we choose, between: what a regulator has published, what a court has found, what a source alleges, and what we have verified ourselves. We do not report an unauthorised firm as a criminal one, and we do not report a Warning List entry as a finding of guilt.

Bylines and sign-off

Every published story carries a named author with a profile page. Publication requires an active editor's approval; the gate is enforced by a database function, so there is no path to publish that skips it. Automated ingestion can fill the queue and nothing else — it cannot publish.

Naming individuals

A story that names a private individual cannot be published until an editor has written the public-interest reasoning. That note is recorded at the time, not reconstructed later, and it is shown on the story. We weigh the seriousness of the conduct, the reader harm being prevented, whether the person holds a position of responsibility, and whether naming adds anything the reader needs.

We do not name children, and we do not name family members or associates who are not themselves part of the conduct.

Right of reply

Where we make a critical claim about a firm or a person, we put the substance to them before publication with a clear deadline, and we record the outcome on the story: response received, declined to comment, or no response by the deadline. A reply received after publication is added to the story, with the date.

Right of reply is free, it is never conditional on anything, and it is never sold. Threats of legal action are logged and do not shorten the reply window or change the standard we apply.

Legal review

Stories flagged as needing legal review cannot be published until that review is recorded against the article. Flags are applied by the author or the editor and, once set, cannot be cleared by the author alone.

Anonymity and confidential sources

We protect the identity of confidential sources. Tips are readable only inside the newsroom, are never published by default, and are purged on a fixed retention clock. We will not reveal a source, and we do not keep material that would identify one longer than the story requires.

Anonymity is granted where a person faces a real risk — to employment, safety or legal position — and never simply to allow an unattributed attack on a named party.

Reporting on people who have been defrauded

We do not ask how much someone lost, and we hold no field for it. We do not publish a person's account without their informed consent, we tell them where their words will appear, and we let them withdraw before publication. We do not use language that implies a person brought fraud on themselves, and we do not photograph or describe anyone in a way that would expose them to repeat targeting — recycled contact lists are a known second wave of harm.

Images, logos and screenshots

A firm's logo is shown only to identify the firm the story is about, at small size, and is fetched through a third-party icon cache so that no reader's browser contacts a warned website. Where no genuine mark exists we show nothing rather than invent one. Screenshots of scam messages are redacted so that no reader's personal details, and no live payment or contact route, are reproduced.

Third-party copyright

Where we reference another publication's work we use a headline, our own two-sentence summary and a link. We do not republish their article, their structure or their phrasing, and we credit the outlet in the visible text, not only in a link.

Updates and versioning

A developing story is updated in place with an "updated" timestamp and a change log entry describing what changed and why. Substantive changes are never made silently. Where an update reverses or materially alters what we previously said, it is treated as a correction, not an update.

Corrections

We correct promptly and visibly. A correction states what was wrong and what changed, appears on the article itself and in the public corrections log. Significant corrections stay visible at the top of the article. We do not delete a story to make an error disappear.

Complaints and takedown requests

Every complaint and takedown request is logged with the date received. Time-limited statutory processes are tracked with a visible clock so deadlines are met. We will correct an inaccuracy, add a reply, or update a story where circumstances have changed — for example where a Warning List entry has been withdrawn. We will not remove accurate, public-interest reporting because it is unwelcome, and we do not accept payment to do so.

Reader comments and moderation

Comments are pre-screened and moderated. We remove content that names a private individual without a public-interest basis, alleges criminality without support, promotes a recovery or claims service, includes contact details or payment routes, or targets a person's protected characteristics. Repeat abuse loses access. Moderation decisions are explained to the commenter and can be appealed through the contact page.

The advice boundary

The Fraud Center publishes information, not advice. We do not assess claims, estimate compensation, or act for readers. Where a reader has lost money we point only to the free routes: their bank, the Financial Ombudsman Service, Action Fraud and the FCA consumer helpline on 0800 111 6768. We will not build a claim-eligibility quiz, a compensation calculator, or a referral route to a representative, and we will decline requests to add one.

If you are unhappy with our response

Write to us first through the contact page, setting out what is inaccurate and what you would like corrected. Complaints are answered by an editor who was not the author. If you remain dissatisfied, you keep every legal route otherwise available to you; nothing in this policy limits your rights.