Fact-Check a Viral Claim with Sources and Confidence

Answer first: Paste the post link or claim and ask Grok Bot to trace it: find the earliest post, check primary sources on the web, list the exact searches it ran, and give a confidence rating with reasons. You get an auditable verdict instead of a vibe.

Test status: the X search, X post-lookup tools and web fetching were confirmed working on 2026-10-16 (searches returned posts with direct links). A full trace of one specific viral claim was not done for this page.

Who it's for

Journalists, community managers and anyone about to repost something surprising.

Steps

1. Copy the post URL, or the exact wording of the claim.
2. Paste the prompt. Ask for the queries it used so you can rerun them.
3. Read the sources it links, not just the verdict.

Copy-paste prompt

Fact-check this claim: "<CLAIM or POST URL>".
1) Search public X (sort by oldest/earliest you can find) for the first appearance and who amplified it.
2) Look for primary sources on the web (official sites, filings, original papers, reputable outlets). Fetch them.
3) Report: verdict (True / Mostly true / Misleading / False / Unverified), confidence (low/med/high) with reasons,
   the 3-5 key sources with links, and the exact search queries you ran.
Flag anything you couldn't open instead of assuming what it says.

Expected result

A verdict, a confidence rating, a short evidence list with links, and a reproducible query log.

Gotchas

• X search has its own time window and returns at most 25 posts per search, so the "earliest post" it finds is the earliest it found. Treat it as a lead, not proof.
• Some sites block automated fetches. A blocked page isn't evidence either way, so it should say "couldn't open".
• "Unverified" is a valid answer. Don't push it to pick a side.
• It won't reply to the post or contact the author.

FAQ

Images or video? It can describe and search context around them, but tracing a manipulated image is harder. Ask it to state the limits.