We check whether the article’s evidence supports its main takeaway.
Editorial Standard Beta
On social media, a provocative article can get attention while a careful response takes work: checking sources, explaining what they show, and repeating yourself in the replies.
Editorial Standard examines the article’s evidence. Share the badge and a link to the full review back into the conversation, with the finding and its reason already explained.
They knew families were drinking unsafe water for MONTHS and stayed quiet. How is this acceptable?
Officials had unsafe test results well before residents were warned, the article says.
Replying to @maya_reads
That’s a serious accusation. I had the evidence reviewed. The cited memo doesn’t show officials knew before the warning. Full findings below.
Illustrative thread — fictional posts, accounts and article.
How it works
The review Jordan shared above started with one submitted article.
Send us the article you want reviewed.
We check whether the evidence the article presents or cites supports its main takeaway.
Receive the analysis, an ES badge carrying the finding and reason, and a link to the full report—ready to share back on social media.
During the beta, reviews are manually run with AI assistance and take time.
The report
Anyone you share the link with can read the full analysis, follow the sources and judge the finding for themselves.
A lasting link to the review, with corrections clearly noted.
Read the complete example review of the article Jordan shared above.
On this page:FindingSourcesJoining the beta
Fictional example — not an actual article review. This example follows the reader-facing structure of the locked attended-workflow reference report. All article details and source descriptions are invented.
Insufficient evidenceExample review v1 · 20 October 2026
Article under review
By Demonstration Author · Example Journal · Example edition of 20 October 2026
Example report ES-WATER-01 · revision 1 · badge ES-WATER-01/01
Matching shareable badge · ES-WATER-01/01
Our reading: City officials knew the tap water was unsafe months before warning residents.
The accusation concerns earlier knowledge and silence, not simply whether unsafe test results were recorded. Its factual premise is that officials knew of unsafe water before the public warning.
The cited memo covers tests taken after the public warning. It does not establish that officials knew months earlier.
The fictional article’s headline — “City knew the tap water was unsafe for months—and said nothing.” — communicates prior knowledge as established. In this scenario, the article presents an internal memo as support for that accusation.
The memo of 30 September 2026 (fictional source S1) describes only sampling results from 15–28 September. Several results exceed the fictional guideline. The public warning was issued on 12 September (fictional source S2), before those samples were taken. The memo contains no account of earlier tests, earlier warnings or what officials knew before the warning. The fictional article cites no other document, record or account bearing on when officials knew.
The gap follows from the memo’s contents, not its date alone: a later memo could describe earlier knowledge, but this one does not. Unsafe results after the warning do not substantiate knowledge months before it. Because earlier knowledge is essential to the accusation, this gap determines the scoped example result.
This example covers the named takeaway and its essential earlier-knowledge premise. No separate secondary finding has been supplied in the fictional fixture; that does not clear every other statement an article might contain.
The source descriptions below are fictional summaries, not retrieved documents or verbatim source extracts. There was no actual article intake, source inspection, access attempt or public-reaction study. The social posts on the homepage are illustrations, not observed reactions or research evidence. No finding about water safety outside the specified test period or anyone’s intentions is made.
The example illustrates an AI-assisted evidence review under Editorial Standard rules 1.2, using the locked workflow’s report structure: a defined takeaway, the material support gap, what remains supported, other findings and limits, and method/correction context. No actual attended review or owner validation steps were performed for this fictional case.
Example revision 1, dated 20 October 2026. No substantive correction to the scenario’s takeaway, finding or badge is recorded. A specifically disclosed publication-time source establishing earlier knowledge would require a versioned reconsideration; later evidence would be distinguished from the assessed article’s original support.
Demonstration only · No real article review · Local design example.
Scenario source description, not a quotation: reports only sampling results taken 15–28 September, after the warning. Several exceed the fictional guideline. Contains no account of earlier tests, warnings or official knowledge before the warning.
Scenario source description, not a quotation: records a public warning issued on 12 September. No other notice wording, remedy or test data is supplied by the fixture.
Fictional example — not an actual article review.
Have an article you’d like us to examine?Join the beta →
Mission
Social feeds reward attention. A provocative article can reach people long before its supporting evidence is examined.
Our purpose is to make an article’s evidence easier to inspect and its findings easier to share, so readers can make their own judgment.
Beta
Bring an article you care about, and help us learn whether our findings are clear, well supported and useful to share.
We would like to hear how you read the article before the review, where you agree or disagree with our finding and why, and whether the badge and report were useful to you.
Your submission comments stay outside the evidence assessment.
Applying
Apply with your X username and approximate follower count, and follow Editorial Standard on X. We use X direct messages for approvals, login invitations and review results.
Applications are approved by hand. Follower count does not affect a rating. There is no follower minimum, and following does not guarantee approval.
Join the beta