Corrections as a Process, Not an Event

Why a policy document is not enough

A written policy tells staff what a correction should look like. It does not, by itself, ensure that every error actually reaches that outcome. In practice, corrections fail not because the policy was wrong but because the process around it broke down: an editor was on leave and no one else picked up the escalation, a story changed twice in one afternoon and no one tracked which version a complaint referred to, or a correction was published without anyone checking whether the corrective wording itself created a new legal risk.

Treating corrections as a process means building in the checkpoints, named owners, and review habits that make the policy work under real newsroom conditions — including at 11pm on a Friday, mid-way through a breaking story, or when the person who wrote the original piece has since left the organisation.

The editor review chain, in practice

Most newsrooms already have an escalation chain on paper. The process layer is about who actually reviews what, and when, rather than simply who is nominally responsible:

• First reviewerThe section or commissioning editor reads the flagged error against the published text within the same working day it is raised, and drafts an initial severity assessment.
• Second reviewer (Tier 2 and above)A senior editor independent of the original commissioning line re-checks the assessment before any correction goes live, to avoid a single editor marking down the severity of their own team's error.
• Legal read (Tier 3, or any defamation/contempt dimension)Legal counsel reviews both the proposed corrective wording and the original text, since the correction itself can create fresh legal exposure if drafted carelessly.
• Final sign-offRecorded against a named individual in the corrections log, with a timestamp — not just a verbal go-ahead in a chat channel that leaves no audit trail.

A severity scale that can move — not just be assigned once

The three-tier severity rating set out in our internal corrections policy guide is a starting classification, not a fixed one. Severity should be re-assessed if new information emerges — for example, a Tier 1 typo in a name turns out to relate to a case of mistaken identity with reputational consequences for a real person, which reclassifies it as Tier 2 or Tier 3 even though the original text looked minor.

Admission of error and defamation implications

Journalists sometimes worry that publishing a correction is equivalent to admitting liability and will be used against the publisher in a defamation claim. In UK practice, the opposite is usually true. A prompt and proportionate correction supports the “responsible journalism” strand of the section 4 public interest defence under the Defamation Act 2013, and is treated by courts as a mitigating factor relevant to damages, since it limits the ongoing harm caused by continued publication of an inaccuracy.

The legal risk is not in correcting — it is in how the correction is worded. A correction that goes further than necessary, that restates the original allegation in a new and equally unsupported form, or that inadvertently implicates a third party, can create a fresh cause of action distinct from the original story. This is why legal read-through of the corrective text itself, not only the original article, is a standard step for Tier 3 corrections.

• ›Correct only what is necessary — do not use the correction as an opportunity to add unrelated new claims.
• ›Avoid repeating the exact wording of a defamatory original claim inside the correction if it can be described accurately without repetition.
• ›Where a complainant has threatened proceedings, route the corrective wording through legal counsel before publication, not after.
• ›Keep a dated record of legal sign-off separate from the editorial sign-off in the corrections log.

Version control for major and running stories

A single archive note works well for a static article corrected once. It is insufficient for a running story — a live court case, an election count, an evolving crisis — where the text may be revised multiple times in a single day as facts develop. Without version control, it becomes difficult to establish exactly what was published at what time, which matters if a complaint refers to “the version I read at 3pm” rather than the current live text.

Archive correction footer conventions

Consistency in how correction footers are written matters almost as much as the correction itself — readers and regulators learn to recognise and trust a publisher's correction format when it follows a consistent convention. A workable footer template:

This template satisfies the core requirement of IPSO Clause 1(ii) — that the reader can see what was wrong and what is now correct — without requiring a bespoke form of words to be drafted under time pressure each time. Where an apology is not warranted (for example, a minor Tier 1 error), the final sentence can be omitted, but the first two should always be present.

Building the process into onboarding and audit

A process is only as reliable as the people running it know it. New editorial hires should walk through a live example of a Tier 2 correction during induction, not simply be handed the policy document to read. Editorial standards meetings should periodically sample recent corrections against the process steps above — did the second reviewer actually review, was legal sign-off recorded, was the footer consistent — rather than only auditing whether a correction happened at all.

This audit habit is also what demonstrates systemic compliance to IPSO or IMPRESS if a publisher's corrections handling is ever formally reviewed: a documented process that is regularly checked against actual practice is stronger evidence of good faith than a policy document alone.

Quick-reference process checklist

• Error flagged and logged with date, source, and article URL.
• First reviewer assessment completed within the same working day.
• Independent second reviewer engaged for Tier 2 and Tier 3.
• Severity re-rating trigger checked if new information or a complaint arrives after initial correction.
• Legal read of the corrective wording itself for any defamation, contempt, or data protection risk.
• Version control / changelog updated for any running story before it is closed out.
• Footer drafted using the consistent correction template.
• Final sign-off recorded against a named individual with a timestamp.
• Correction sampled in the next editorial standards meeting for process — not just outcome — audit.

Frequently asked questions