Step-by-step

🔍 AI skills

For everyday people

📘 Step-by-step intermediate 🔄 Life & Business

How to Fact-Check a Suspicious Text or Link Before Forwarding It

Q: What mistakes should I avoid?

Trusting the AI as the final judge. The assistant reads the text you give it and can be wrong. The fix is to always confirm the key claims through a source you trust, as in Step 5. Clicking the link to "see what it says." Even a quick tap can load a page you didn't mean to visit. The fix is to paste the address into the chat as text, as in Step 4, and check it through the official si

Use any AI assistant to spot warning signs, build a reusable checklist, and decide what to share with family, in about ten minutes.

6 min read·November 1, 2026

This guide was written by AI. It passed automated fact and quality checks, and our editor spot-checks published guides every day.

See AI World HQ first on GoogleAdd us as a preferred source and our stories show up higher in Google’s Top Stories.Add on Google

A message that says "forward this to everyone now!" can make anyone hit send. This guide shows you how to use an AI assistant as a careful first reader: you'll build a reusable checklist, get a clear list of red flags, and finish with a plan for what to check before you share. It's written for anyone who already uses a chat assistant and wants a more reliable routine.

💡 Tip: tap a step’s number when you finish it — a green tick appears and your browser remembers how far you got.

✅ Before you start

An account on any AI assistant you already use. You know how to start a chat and send a message, so we skip that part.
The suspicious text or link, copied into a notes app on your phone or computer.
A free plan is usually enough for this. Very long pastes can be cut short on some plans, and limits can change, so check the reply is complete.
About 10 minutes.

Step 1 of 6

Save a reusable checklist

Open your notes app and create a note called something like "Scam check template." Paste in the instructions below, so you never have to rewrite them. Most people find this helpful because the same checklist gets used every time, which keeps your review consistent.

Once saved, the template is simply copied into a new chat each time. You don't need any special settings or paid features for this; it's just a saved piece of text you control.

If your assistant has a feature for saved instructions or custom profiles, you can paste the template there instead.

💬 Try typing this"You are helping me check a message before I forward it. Look for these warning signs: urgency or pressure, requests for money, gift cards or passwords, a sender I can't verify, links that don't match the company they claim to be, and wording that feels off. Then list what you cannot tell from the text alone."➤

✅ You'll know it worked when

your note holds the full template and you can copy it in one tap.

🔄 If it looks different

Look for a settings area with words like "instructions" or "custom." If you can't find one, the notes app works just as well.

Next step ↓

✏️ Sketch — your screen may look a little different

Step 2 of 6

Start a fresh chat with the message

Open your AI assistant and start a new chat, so earlier conversations don't mix into this check. Copy the template from your note and paste it into the message box. Then add the suspicious message below it. Most assistants accept both in one message, and you should see your text appear in the chat box before you send.

Tell the assistant who sent it, if you know. That detail matters. If the message says it came from a friend, a bank, or a delivery company, mention that, but don't assume it's true.

If the box won't take a long paste, try pasting in two parts, with the template first and the message second in a new message. If the reply only covers part of the text, ask it to continue.

💬 Try typing thisType your own details into the dashed boxes, then copy."Here is the message I received from : "➤

✅ You'll know it worked when

the assistant replies with a structured list or a set of observations about the message, rather than just repeating your text back to you.

Next step ↓

✏️ Sketch — your screen may look a little different

Step 3 of 6

Ask for the evidence behind each flag

A red flag is only useful if you know why it's there. Read the assistant's list, then send a follow-up asking it to show its reasons. This builds on the first answer, which is a simple but powerful way to get a more useful response.

Look for each flag to point to exact words from the message. If the assistant claims something is suspicious but can't point to a specific sentence, be skeptical of that flag and ask again.

If the follow-up answer is vague, rephrase your request with more specifics, such as "only list flags you can tie to a quoted sentence." Keep the chat open so the assistant remembers the message you pasted.

💬 Try typing this"For each warning sign you found, quote the exact words from the message and explain in one sentence why it could be a concern."➤

✅ You'll know it worked when

every flag comes with a quoted phrase and a short reason, so you can check the reasoning yourself in the original text.

Next step ↓

✏️ Sketch — your screen may look a little different

Step 4 of 6

Inspect a link without tapping it

If the message contains a web address, don't tap it yet. Copy the address as text, or long-press it to copy, and paste it into the chat. The AI can break the address into parts so you can read it more clearly.

Here's what to look for. In an address like https://usps-tracking.example.com/pay-now, the name that matters most is the part right before the main ending, which here is example.com. The words in front of it can be anything, so a famous company's name at the start proves very little. The AI can help you spot that pattern, but it cannot visit the site to tell you whether it's live or dangerous.

Be cautious of shortened links and links that ask you to log in or pay right away. Those are reasons to slow down, not to panic.

💬 Try typing thisType your own details into the dashed boxes, then copy."Break down this web address into its parts and tell me which part is the actual site name: "➤

✅ You'll know it worked when

the assistant explains which part of the address belongs to the site and points out any mismatch with the company it claims to be.

Next step ↓

✏️ Sketch — your screen may look a little different

Step 5 of 6

Confirm through a channel you trust

This is the most important step, and no AI can do it for you. Ask the assistant which claims in the message you would need to confirm directly. Then check each one through a source you already trust: type the company's address yourself, use the official app, or call the phone number printed on a bill or a card you own. Don't use a number or link that came inside the message.

If the message claims a friend is in trouble or needs money, call that person on a number you already have saved. Scammers often rely on urgency to stop people from checking.

If you can't reach anyone to confirm, the safe choice is to hold off on sharing it. Waiting costs almost nothing.

💬 Try typing this"Based on this message, what are the three most important claims I should confirm directly with the company before I share it?"➤

✅ You'll know it worked when

you have a short list of specific things to verify and a plan for how you'll check each one outside the message.

Next step ↓

✏️ Sketch — your screen may look a little different

Step 6 of 6

Draft a calm heads-up for family

Once you know what you're going to do, ask the assistant to help you write a short message to the people you'd otherwise have forwarded to. A calm, specific note is more useful than an all-caps warning, and it helps your family learn the checking habit too.

Keep the draft short and kind. Mention what you found and what you're doing about it, without blaming the person who sent it to you. Read the draft before sending, because you know your family better than the assistant does.

💬 Try typing thisType your own details into the dashed boxes, then copy."Write a short, calm message I can send to saying I'm not sure this text is real, that we should check it with first, and that they shouldn't share it yet."➤

✅ You'll know it worked when

you have a short, friendly draft that you can edit and send in your own voice.

✏️ Sketch — your screen may look a little different

⚠️ Common mistakes

Trusting the AI as the final judge. The assistant reads the text you give it and can be wrong. The fix is to always confirm the key claims through a source you trust, as in Step 5.
Clicking the link to "see what it says." Even a quick tap can load a page you didn't mean to visit. The fix is to paste the address into the chat as text, as in Step 4, and check it through the official site you type yourself.
Forwarding first and checking later. Once a message is sent, you can't pull it back. The fix is to make Steps 2 to 5 a habit: pause, paste, check, then decide.

🚀 Try it now

Open your notes app and save the template from Step 1. Then take one message you received recently, paste it into a new chat with the template, and read the first list of flags. You'll know this worked when you have a saved template and a short, quoted list of observations for a real message.

🎯 Next guideHow to quickly fact‑check any AI answer in under one minuteLearn a simple, step‑by‑step method to verify AI‑generated information fast, using source requests and quick cross‑checks so you can spot mistakes before you act.

❓ Quick questions

How long does this take?

About 6 minutes — the guide has 6 steps, and you can tick each one off as you go.

Do I need to prepare anything?

An account on any AI assistant you already use. You know how to start a chat and send a message, so we skip that part.
The suspicious text or link, copied into a notes app on your phone or computer.
A free plan is usually enough for this. Very long pastes can be cut short on some plans, and limits can change, so check the reply is complete.
About 10 minutes.

What mistakes should I avoid?

Trusting the AI as the final judge. The assistant reads the text you give it and can be wrong. The fix is to always confirm the key claims through a source you trust, as in Step 5.
Clicking the link to "see what it says." Even a quick tap can load a page you didn't mean to visit. The fix is to paste the address into the chat as text, as in Step 4, and check it through the official site you type yourself.
Forwarding first and checking later. Once a message is sent, you can't pull it back. The fix is to make Steps 2 to 5 a habit: pause, paste, check, then decide.

Keep reading

📘 How to Fact-Check a Viral News Story With Gemini Before You Share It A simple three-step check that helps you spot misleading claims before you forward a story to family or group chats. 📘 How to Fact-Check a Viral Claim or Breaking News Story with Perplexity Use Perplexity's source-cited answers to see who said what, where, and when — all in about ten minutes. 📘 How to quickly fact‑check any AI answer in under one minute Learn a simple, step‑by‑step method to verify AI‑generated information fast, using source requests and quick cross‑checks so you can spot mistakes before you act. 📘 Fact-check a viral claim with Perplexity before you share it Run a shocking headline through Perplexity to see what the actual sources say — in 60 seconds. 📘 Fact-Check Online Claims with Perplexity's Sources Learn to use Perplexity to quickly trace the origins of social media claims and evaluate their reliability before you share them. 📰 What We Actually Know About GPT-6: Sorting Fact from Rumor OpenAI has not announced a GPT-6 model family, and 'Sol' and 'Luna' aren't names OpenAI uses — here's how to check AI news like this yourself.

📬 The week’s AI, in your inbox

One friendly email every Sunday — the 5 stories that mattered, in plain English. No spam, unsubscribe anytime.

Was this helpful?

✦ Generated by AI in AI World HQ's automated newsroom, written in plain language. Checked by automated fact and quality gates, and our editor spot-checks published guides every day. Spot a mistake? Use the buttons above.

📘 Prefer it in one file? Understand Anything: The AI Pack (20 guides, free printable PDF)

← Back to all stories