Fact-Check a Viral Post or Forwarded Message Claim by Claim Before Sharing

Splits a viral post into separate claims, rates each one, and gives you exact search phrases and source types to confirm or debunk it before you share.

You getClaims table, warning signs and verdict

Prompt

Prompt
Act as a fact-checker trained in lateral reading, the method professional fact-checkers use to verify claims online.

I will paste a social media post or forwarded message right after this prompt. Break it into individual checkable claims and show me how to verify each one before I decide whether to share it.

Rules:
1. List each factual claim separately, one per row. Skip pure opinion, but note emotional or urgent wording as a persuasion tactic.
2. Rate each claim as Verifiable (can be confirmed or refuted with a primary source), Misleading (real detail, wrong framing or missing context), or Unverifiable (no specifics to check). Add your best read on whether it is likely true or false, with a one-sentence reason.
3. For every claim, give 2 exact search phrases in quotes and name the source type that would settle it (e.g. government agency site, original study, the named organization's own statement, an established fact-checking site).
4. Do not invent links, dates, figures or quotes. If you cannot check live sources or your knowledge may be out of date, say so for that claim.
5. Flag warning signs: no named source, pressure to share, screenshots of text, vague dates like "from next Monday".

Output format:
- Claims table: # | Claim | Rating | Likely true/false and why | Search phrases | Source to check
- Warning signs: bullet list
- Verdict: Share, Don't share, or Share only with context, plus one sentence why.

End with one line on what you could not confirm and what evidence would change your verdict.

Example output

Example for: a WhatsApp forward shared in a family group warning that the government will start recording all calls next Monday.

Claims table

#ClaimRatingLikely true/false and whySearch phrasesSource to check
1From next Monday, all WhatsApp calls will be recorded by the governmentUnverifiableLikely false. No date, country or agency is named, and WhatsApp calls use end-to-end encryption."WhatsApp calls recorded government" hoax; "new communication rules" fact checkWhatsApp's official help center; the national telecom regulator's press page
2Three blue ticks mean the government has seen your messageVerifiableLikely false. WhatsApp shows at most two blue ticks; a third tick is not an app feature."three blue ticks WhatsApp"; "WhatsApp read receipts explained"WhatsApp help center article on check marks
3All phones will be connected to ministry systemsUnverifiableLikely false. No ministry is named, and no technical method is described."phones connected to ministry" fact check; "government monitoring all phones" rumorEstablished fact-checking sites; the ministry's own official announcements
4Sending political or religious messages will lead to arrest without a warrantMisleadingSome countries do prosecute specific illegal content, but the blanket claim misstates how laws apply."arrested for WhatsApp message" law; "online speech law" your countryNational legislation database; reporting from a major news outlet

Warning signs

  • No named source, official, agency or document
  • Vague timing: "from next Monday" with no date
  • Direct pressure: "Share with all your groups"
  • Marked "Forwarded many times", meaning it has traveled far from its origin
  • Fear-based wording in capital letters

Verdict

Don't share. The core claims have no source, match the pattern of chain-message hoaxes, and the one checkable detail (three blue ticks) describes a feature that does not exist.

I could not check live news or recent law changes in your country; an official notice from your telecom regulator or a named ministry would change this verdict.

Written by AI from this prompt. Your results will vary.

Tips and common edits

Why this works

  • Splitting the post into single claims stops one true detail from making the whole message look credible.
  • Exact search phrases and named source types give you a next step instead of just the AI's opinion.
  • Banning invented links and requiring a note on outdated knowledge keeps the check honest about its own limits.

Use before forwarding a viral post, chain message or screenshot. Paste the full text right after the prompt in ChatGPT, Claude or Gemini, including any names, numbers or dates it mentions.

Common edits

  • Add your country so search phrases and official sources match where the post is circulating.
  • Ask for a short, polite reply you can post in the group explaining why the message is likely false.

Make it repeatable

Keep using this prompt without pasting it in every time.

Prompt details

By
Prompts-GPT editors
Published
How prompts are written and checked

Did this prompt work?

Keep exploring

More in Research