Social Media Verification Checklist — Twelve Checks Before You Use a Post
A recycled photograph from another country and another year is the single most common way a newsroom gets caught out on a breaking story. The checks that catch it are known, take minutes, and are exactly the ones skipped under deadline.
Last reviewed: Next review due:
Social Media Verification Checklist
Answer Yes or No for each verification step before using social media content in a story.
1.Have I traced the content to its original source (not a screenshot of a screenshot)?
2.Have I checked the account's history (created when, posting frequency, language patterns)?
3.Does the account have a verified blue tick or established credible footprint?
4.Have I cross-referenced the claim with a second independent source?
5.Have I run a reverse image search (Google Lens, TinEye, Yandex) on any images?
6.Have I checked timestamps against the claimed event time (and timezone)?
7.Have I geolocated images/video against known landmarks, signage, language?
8.Have I checked image metadata where available (EXIF — but note social platforms strip it)?
9.Have I checked video for editing signs (jumps, audio glitches, unnatural shadows)?
10.If the content is sensitive (atrocity, death, identifiable victims), have I assessed it ethically before publishing?
11.Do I have permission from the original poster, OR is the content clearly newsworthy enough to use under fair dealing?
12.Have I screen-captured the content with URL and timestamp visible (in case the post is deleted)?
How it works
- Twelve yes/no items: tracing the content to the original post rather than a screenshot; checking account age and posting patterns; treating a blue tick as unproven identity; finding a second independent source; running a reverse image search; checking timestamps against the claimed event and timezone; geolocating against landmarks and signage; checking EXIF on original files; looking for signs of video editing; assessing the ethics of sensitive content against IPSO Clauses 4 and 9; securing permission or relying on fair dealing under s.30 CDPA 1988; and archiving the post with URL and timestamp.
- All twelve must be answered before any result appears.
- The score is the number of Yes answers: 10-12 gives “Strong” with a green badge reading “Verification steps complete — now do your legal and editorial checks”, 7-9 gives “Moderate — Hold and verify further”, and 0-6 gives “Weak — Do not publish”.
- Each No prints a fixed tip naming the tool to reach for: TinEye, Google Lens, Yandex Images, InVID/WeVerify, SunCalc, ExifTool, FotoForensics, the Wayback Machine and full-page capture tools.
When to use it
- On a breaking story where a viral image or clip is the only evidence you have.
- When user-generated footage is offered to your desk and you need a defensible record of what was checked.
- Before republishing content about an identifiable victim, where the ethics and copyright items matter as much as the verification ones.
- Training a newsroom on the difference between having seen a post and having verified it.
What it does not do
- The badge is produced by counting your own ticks. The tool never looks at the post, the image, the account or the metadata — it cannot tell a real check from a claimed one. Reaching “Strong” does not mean publish: the tool deliberately stops at telling you the verification steps are done and that legal and editorial checks still follow.
- The items are unweighted, so 10 out of 12 says Publish even when the two you skipped are the reverse image search and the geolocation — the two checks most likely to catch a recycled image.
- Two of the twelve are not verification at all: the sensitive-content ethics item and the permission/fair-dealing item. Ticking them raises the “verification confidence” number exactly as much as running a reverse image search does.
- It archives nothing for you. Item 12 asks whether you have captured the post; if it is deleted before you act on that prompt, there is no recovery.
- Synthetic and AI-generated media are out of scope. “Check video for editing signs” is an instruction to look, not a forensic test, and none of the named tools reliably detect current generative video.
More tools
This is one of the free tools on UK JournoHub. See the full tools index for the rest.
Related guides
Primary sources
- CDPA 1988 s.30 — fair dealing for reporting current events— legislation.gov.uk