Step-by-step

🛡️ AI skills

For everyday people

📘 Step-by-step beginner 🏠 Everyday life

How to Spot a Phishing Email or Scam Text Before You Click

Paste any suspicious message into an AI assistant with a fraud-detector prompt and it will flag the red flags your eyes missed

6 min read·August 12, 2026

This guide was written by AI. It passed automated fact and quality checks; no human editor reviewed it.

That pang of "wait, something feels off" you get when a strange email or text arrives — the AI can confirm it (or calm you down) in about 30 seconds. This guide shows you exactly what to type, and how to read the answer so you don't miss the warning signs. It works with ChatGPT, Gemini, Claude, Copilot, Le Chat, DeepSeek, or any other assistant you already use. One honest note up front: an AI can spot a lot, but it can't actually click the link for you or verify whether a sender is real — the final "yes" or "no" is still your call.

💡 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 (free versions work fine — ChatGPT, Gemini, Claude, Microsoft Copilot, Le Chat, or DeepSeek all do this).
The suspicious message itself, opened on your screen but with no links tapped or replies sent.
Roughly 2 minutes.

Step 1 of 6

Read the message calmly before doing anything

Open the suspicious email or text, and do exactly one thing: read it top to bottom with your finger off the mouse. Glance for three things — a sender or number you don't recognize, a tone that pushes you to act NOW, and any link you weren't expecting. You'll see the message exactly as you normally would; nothing changes yet.

💬 Exampleif it's "Your package is delayed, confirm here: bit.ly/x9zq", just read the words. Don't tap the link.

Reading is the goal here.

🔄 If it looks different

in some messaging apps long-pressing a link shows a small URL preview — that's actually useful, but still don't open it.

✅ You'll know it worked when

you can describe the message in one short sentence without having clicked anything.

Next step ↓

Read the message calmly before doing anything

Step 2 of 6

Open any AI assistant you already use

On your phone or laptop, open the AI assistant app or website where you already have an account. No special "fraud tool" is needed — every general-purpose assistant can do this job. You'll see an empty chat box, usually labeled "Message", "Ask anything", or similar. If you're stuck, start a "New chat".

🔄 If it looks different

every assistant has its own empty state, but the chat box is almost always in the middle of the screen on desktop or at the bottom on mobile.

✅ You'll know it worked when

you see a blinking cursor and a send button, ready for you to type.

Next step ↓

Step 3 of 6

Paste in a "fraud detector" prompt first

Before pasting the suspicious message, give the AI a one-line role instruction. Type it into the chat box and hit send. The AI will respond with a short confirmation — usually something like "Sure, paste it in" — that primes it to focus on fraud patterns instead of generic advice. This single line is what turns a normal chat into a fraud check.

💬 Try typing this"Act as a fraud detector. Read the message I'm pasting next and list every red flag you can see, then tell me clearly whether it's safe to reply, click any link, or open any attachment. If you're not sure, say so."➤

🔄 If it looks different

some assistants reply with a longer setup paragraph — that's fine; what matters is that it now knows to focus on scams.

✅ You'll know it worked when

the AI answers something like "Got it — paste the message" or asks a clarifying question.

Next step ↓

✏️ Sketch — your screen may look a little different

Step 4 of 6

Paste the full suspicious message, then send

Highlight the entire message — sender name, subject line, body, and the link text — but leave out your own name, address, or any account numbers if they appear, since the AI doesn't need them to spot trouble. Paste it into the chat box and hit send. The AI now has the raw text and starts analyzing it.

💬 Exampleof what to paste: "From: Amazon Support <[email protected]> — Dear customer, your account has been suspended due to unusual activity. Click here to verify: hhtp://amaz0n-secure.example.com/login — please respond within 24 hours."

🔄 If it looks different

if the message is unusually long (a fake invoice, say), you can paste a trimmed version — but keep the opening sentence and any links exactly as written, because that's where the red flags usually live.

✅ You'll know it worked when

the AI starts replying with a bulleted list of specific things it spotted — usually 4 to 8 items.

Next step ↓

✏️ Sketch — your screen may look a little different

Step 5 of 6

Read the red flags and trust the obvious ones

The AI's reply is usually a checklist: things like "the sender domain doesn't match Amazon," "the URL uses a zero instead of an 'o'," "the message creates urgency with a 24-hour deadline," or "the greeting is generic (no real name)." Read each one and ask yourself, "Does this match what I'm seeing on my screen?" You'll start to see a pattern you can recognize next time without any help.

💬 Exampleof an AI answer: "Red flags: sender domain is amaz0n-support.example.com, not amazon.com; the link is 'hhtp' (typo) and uses a zero in 'amaz0n'; urgency tactic ('within 24 hours'); generic greeting. Verdict: do NOT click, do NOT reply. Report and delete."

That's the same meaning. If it ever says "I'm not certain", take that seriously and lean toward treating the message as a scam.

🔄 If it looks different

the AI might use softer words ("likely a phishing attempt", "this looks suspicious").

✅ You'll know it worked when

you can name at least three specific reasons the message is suspicious and explain them out loud.

Next step ↓

✏️ Sketch — your screen may look a little different

Step 6 of 6

Decide, then report and delete

Once the AI gives its verdict, decide based on the strength of the red flags. If even one major red flag is present — a sketchy sender domain, pressure to act NOW, or a request for passwords or payment — don't click and don't reply. Use the email or messaging app's built-in "Report phishing" or "Report junk" option if you can find it, then delete the message. The dangerous message is gone from your screen and flagged to the platform.

💬 Exampleof a good follow-up question to the AI: "Can you explain in one sentence what an 'urgency tactic' means, so I know what to look for next time?"

If you can't find the report option, deleting is good enough — what matters is that you don't reply.

🔄 If it looks different

Gmail, Outlook, Apple Mail, and most phone carriers all have a Report button, usually under a "⋯" three-dot menu or near "Mark as" / "Move to".

✅ You'll know it worked when

the message is no longer in your inbox, you've flagged it to your email or carrier, and you haven't tapped the link.

⚠️ Common mistakes

Replying to the scammer just to ask "is this real?" — replying confirms your address is live, and more scam messages follow. Delete and report instead.
Trusting the AI without reading the explanation — AI can miss things. If the message "feels" off but the AI says it's fine, trust your gut and don't click. The AI is a second opinion, not a final one.
Pasting personal details along with the message — your name, address, and account numbers aren't needed to spot red flags; trim them out before pasting, because anything you send into a chat is processed by that service.

🚀 Try it now

The next time an email or text feels off, copy it into your AI assistant with the prompt from Step 3, paste the full message from Step 4, and read the red flags from Step 5 before you tap anything. The whole check takes about half a minute.

🎯 Next guideHow to Spot AI-Generated Scam Messages Before You Click or ReplyA practical five-minute checklist for checking suspicious texts, emails and DMs with your usual AI assistant

❓ 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 (free versions work fine — ChatGPT, Gemini, Claude, Microsoft Copilot, Le Chat, or DeepSeek all do this).
The suspicious message itself, opened on your screen but with no links tapped or replies sent.
Roughly 2 minutes.

What mistakes should I avoid?

Replying to the scammer just to ask "is this real?" — replying confirms your address is live, and more scam messages follow. Delete and report instead.
Trusting the AI without reading the explanation — AI can miss things. If the message "feels" off but the AI says it's fine, trust your gut and don't click. The AI is a second opinion, not a final one.
Pasting personal details along with the message — your name, address, and account numbers aren't needed to spot red flags; trim them out before pasting, because anything you send into a chat is processed by that service.

Keep reading

📘 How to check if a suspicious message is a scam by pasting it into an AI A 5-minute way to get a fast second opinion on a sketchy email or text before you click anything 📘 How to Spot an AI Scam Call, Fake Email, or Deepfake Voice Today Three AI-powered scams happening right now and the one-second checks that catch each one 📘 How to Spot AI-Generated Scam Messages Before You Click or Reply A practical five-minute checklist for checking suspicious texts, emails and DMs with your usual AI assistant 📘 How to Spot AI Scams, Deepfakes, and Fake Content Learn the practical red flags to protect your family and business from realistic AI-generated tricks in our digital world. 📘 Turn a Messy Email Into a Clear, Polite Reply With AI Paste that confusing message into any AI assistant and walk away with a confident, professional reply in minutes. 📘 How to Ask GitHub Copilot to Explain a Confusing Error Message A beginner's way to turn cryptic red text into a clear, friendly to-do list

Was this helpful?

✦ Generated by AI in AI World HQ's automated newsroom, written in plain language. Checked by automated fact and quality gates — no human editor reviewed this guide. Spot a mistake? Use the buttons above.

📘 Prefer it in one file? Scams, Deepfakes and Privacy: The AI Safety Pack (9 guides, free printable PDF)

← Back to all stories