Reality Check Quickstart: Verify Before You React
A calm first pass through suspicious messages, calls, images, sellers, and viral claims before you trust, pay, repost, or panic.

A suspicious message, call, image, listing, or viral claim can ask you to decide before you have enough information. Take a breath. Preserve what you received, separate the claim from the route that delivered it, and check it through a source the sender did not supply.
The human pattern underneath
Urgency, fear, affection, embarrassment, and outrage can all make a request feel more urgent than the evidence warrants. A familiar name, a polished page, or a convincing voice may make it feel personal. None of those details alone proves who is asking or whether the story is true.
You do not need to distrust every message. Notice when someone asks for money, access, secrecy, or a quick public reaction before you can check the facts. A genuine problem may still need attention; the aim is to choose a safer route to it.
Quick facts
| Question | Practical answer |
|---|---|
| Level | Beginner |
| Time | 10 minutes for a first pass; difficult claims can take longer. |
| First move | Pause before clicking, paying, reposting, downloading, replying, sharing a code, or agreeing to secrecy. |
| Stronger evidence | Check a known contact, an official source, or the original context independently. Caller ID, screenshots, and a single detector result are not proof. |
| Escalate when | Money, credentials, account access, intimate images, minors, threats, or legal concerns are involved. |
Plain definitions
| Term | Plain meaning |
|---|---|
| Known channel | A contact path you trusted before the suspicious request, such as a saved phone number, official app, bookmarked site, or in-person conversation. |
| Pressure clock | A deadline used to push you into acting before checking; a deadline alone does not prove a request is false. |
| Verification loop | A short sequence of preserving the request, checking its source and context, and deciding what to do next. |
The practical workflow
| Step | What to do |
|---|---|
| Pause | Stop the conversation or leave the page. Do not click links, open attachments, pay, share codes, or repost while you check. |
| Save | Keep the original message where possible and note the date, sender, requested action, and any relevant transaction details. Take a screenshot without opening a suspicious attachment. Keep the record private. |
| Separate | Write down the specific claim and what it asks you to do. A real organization or a familiar account can be impersonated; an authentic image can be presented with a false caption. |
| Verify | Contact the person or organization through a known route, or find the original source and context of a public claim. Do not use the number or link in the suspicious request. |
| Decide | Continue through the verified route if the specific request is confirmed; stop and report if it fails; leave it unresolved if you cannot check it. Escalate promptly if harm or access is already involved. |
A known route is useful only if it still reaches the right person. If a saved account may be compromised or a number has changed, use another independently confirmed route. For an unexpected bank call, hang up and start a new contact yourself. The FTC recommends using the official app, a recent statement, or the number on your card. Do not let the caller transfer you or guide your search for a number. For a family emergency, call the relative on a number you know or check with someone else close to them; a familiar voice alone is not enough.
If you do not already have a contact, find the organization independently through a source you trust and cross-check it against existing records. A sponsored search result or a second message from the same account is not an independent check. The lookalike domain guide can help you find the right site.
Ask an open question first, such as “What do your records show?” Then check the exact action, not merely whether the person or account exists. If nobody can confirm it, an unanswered call is not proof that the emergency is false. Seek local emergency help if there is an immediate risk to someone’s safety.
Match the check to the claim
- A request to pay or change account details: Ask an authorized person through a previously established contact whether they made this specific request. Confirm the recipient, amount, and full payment destination through the organization’s secure process before sending anything. Confirmation that an invoice exists does not confirm a change of destination. See Known-Channel Callback .
- An image, screenshot, or viral claim: Look for the earliest available source, date, location, and surrounding context, then check whether they support the caption. Seek reliable independent confirmation; several accounts repeating the same unverified post do not count as separate evidence. A reverse-image search can find earlier appearances, but a match does not establish what an image depicts or prove it is unaltered; no match does not prove it is new or AI-generated. Use the viral claim check and reverse-image workflow .
- A seller or listing: Check the seller and the item independently, and understand the platform’s payment and dispute process before paying. Photos and positive-looking reviews are clues, not guarantees. Start with the marketplace seller check .
A grounded example
A text says a package is held for a small fee. Instead of tapping the link, save the message and open the delivery company’s app or a bookmarked site. Check a tracking number from your order or contact the carrier through its established support route. Ask whether this shipment has a fee and how it can be paid. If the carrier cannot confirm it, do not pay through the text; an empty tracking page alone may not settle every delivery issue. The FTC’s package-text guidance also suggests checking shipping information in the account where you placed the order.
If you already acted
Do not wait to finish an investigation. If you sent money, contact the bank, card issuer, payment app, or transfer service immediately through a known channel; explain what happened and ask whether the transaction can be stopped or reversed. Recovery is not guaranteed. Keep receipts and transaction details. See what to do after sending money .
If you shared a password or a login code, use the account provider’s official recovery process, change the affected password, and review recent sign-ins, recovery details, and account access. If you reused that password, change it on those accounts too. Turn on two-factor authentication if available, and review any existing authentication methods for changes you did not make. If you granted remote access, end the session and contact the provider or a trusted technician through an independently chosen route. If you suspect your device is compromised, use a trusted device to seek help and recover accounts. If you opened an attachment, seek device guidance from the relevant provider or a qualified support person. The account access guide has more steps.
For immediate physical danger, contact local emergency services. For threats, intimate material, or a child involved, seek appropriate local support and reporting help. Do not forward intimate images or collect more harmful material to prove a point.
Keep the decision reversible
Make a short private note with the contact path, times, the request, what you did, and the result of any independent check. Save relevant links, handles, screenshots, and payment details without opening suspicious files. Keep passwords, one-time codes, and recovery codes out of the note. Share only what the bank, platform, or other support channel needs, and protect sensitive information. The verification notes guide offers a simple record.
Avoid publicly accusing someone while the facts are uncertain. If a check fails, stop using the suspect channel and report or block as appropriate. If you cannot verify the claim, it is fine to say “I don’t know yet” and hold off on the payment or repost. A reporting route depends on where you live and what happened; the reporting map is a starting point.
Common mistakes
- Treating caller ID, a screenshot, a familiar voice, or a confident tone as identity proof.
- Calling a number supplied in the message or relying on a top search result without checking it independently.
- Reading a one-time login code to an unexpected caller or granting remote access to fix a supposed emergency.
- Treating one AI detector result, or the absence of a warning, as a final verdict.
- Letting embarrassment or pressure to keep a secret prevent a conversation with someone trusted.
Try this next
- Practice a known-channel callback and agree on a family safe-word routine before an urgent call arrives.
- Make a private evidence note for a request you are still unsure about.
- Keep the Reality Check Desk glossary and guidebook shelf handy for a more specific check.
Related Fondsites path
Official references
- FTC: How to handle unexpected calls that claim your money is at risk — independent contact and account-code precautions.
- FTC: Scammers use AI to enhance their family emergency schemes — verify a familiar-sounding voice through a known number.
- FTC: Think that text message is from USPS? It could be a scam — check package and tracking information independently.
- FTC: What to do if you were scammed — payment and account recovery steps, with U.S. reporting options.
- CISA: Recognize and report phishing — suspicious messages and reporting.
- FBI Internet Crime Complaint Center (IC3) — U.S. internet crime reporting.
Written By
JJ Ben-Joseph
Founder and CEO · TensorSpace
Founder and CEO of TensorSpace and the editor of Fondsites: an engineer working across software, AI, cybersecurity, and biosecurity who builds the site's games and tools.