Use AI to Check Suspicious Messages & Links

You now know the scam patterns. But in the moment, when a message says your account will be frozen in 30 minutes, it is hard to think clearly. This is where AI becomes your calm second opinion. In seconds, you can paste a suspicious message, describe a phone call, or check a link, and get a clear risk assessment plus a safe next step. This lesson gives you exact prompts to keep on hand.

What You'll Learn

• How to have AI analyze a suspicious message for scam signals
• How to check links and web addresses safely without clicking
• How to verify whether a company or app is legitimate
• The limits of AI here and how to stay safe overall

The Scam-Check Prompt

Keep this prompt saved on your phone. When any suspicious message arrives, paste it (remove your personal details first) and ask:

"You are a fraud-detection expert. I received this message and I am not sure if it is a scam. Analyze it for red flags, tell me how risky it is on a scale of low, medium, or high, explain exactly which parts are suspicious, and tell me the safe next step. Do not assume it is real. Here is the message: [PASTE MESSAGE]."

AI is excellent at this because it has seen countless scam patterns. It will flag the urgency, the mismatched sender, the request for a code, or the too-good-to-be-true offer, and it will tell you not to click and to verify through the official channel instead.

Important: never include your own OTP, password, or full card number even when checking a scam. Replace them with "[REMOVED]". The scam signals live in the wording and structure, not in your secrets.

Check a Link Without Clicking

Suspicious links are the delivery method for most phishing. Never click to "see where it goes." Instead, copy the link text and ask AI to inspect it:

"I received this link in a message claiming to be from my bank. Without me clicking it, analyze the web address itself. Does the domain match the real bank, or is it a look-alike? Point out anything suspicious like extra words, misspellings, unusual endings, or a different domain hiding the real destination. Link: [PASTE LINK TEXT]."

AI can spot tricks the eye misses, like paypa1.com (a number one instead of an L), secure-bank-login.xyz (real bank name buried in a fake domain), or a familiar name followed by extra words before the real domain. The genuine site's address is short and matches the brand exactly. When unsure, ignore the link entirely and type the official web address yourself or open the app.

For an extra layer, use a search-first tool:

"Search for the official website of [company name] and tell me the exact correct web address, with a citation, so I can compare it to a link I received."

Verify a Company or App Is Legitimate

Before trusting a new wallet, investment app, or "lender," check it. Perplexity shines here because it searches live and cites sources:

"I am considering using an app called [name] for payments/investing. Search for information about whether it is legitimate: is it a registered/regulated company, are there scam reports or complaints, how long has it existed, and what do reviews say? Give me a balanced summary with sources, and flag any warning signs."

Look especially for: regulatory registration in your country, a real physical presence, a track record longer than a few months, and independent reviews rather than only glowing testimonials on the app's own site. A brand-new app promising guaranteed high returns with no regulatory footprint is a major warning sign.

Reverse the Pressure: Ask AI for the Safe Script

If a scammer has you on the phone or chat and you feel pressured, AI can hand you the words to safely disengage:

"Someone is pressuring me on the phone claiming to be my bank's fraud team and wants me to move money. Give me a short, firm script to end the call safely, and tell me exactly how to independently verify whether there is a real problem with my account."

Having a script ready removes the panic. The safe answer is almost always: hang up, then call the official number on your card yourself.

The Limits: AI Is a Second Opinion, Not the Final Word

AI is a powerful helper, but keep perspective:

• AI can be wrong. A clever scam might read as low-risk, or a legitimate but odd message might read as high-risk. Treat the assessment as a strong hint, not proof.
• AI does not know your account. It cannot confirm whether there really is a problem; only your bank can, through official channels.
• The final safety rule stands: when in doubt, do not pay and do not share secrets. Verify independently. AI helps you decide, but your caution is the real protection.

Try It Now: Run Two Real Checks

First, find a suspicious message you have received (most people have several in spam). Remove any personal details and run the scam-check prompt above in ChatGPT or Claude. Note which red flags it identifies.

Second, take a well-known company you use and ask Perplexity: "What is the exact official website address of [company], with a citation?" Compare it to any links you have received claiming to be from them. Save both prompts somewhere you can reach them instantly, because the best time to have them is before the scary message arrives.

Key Takeaways

• Keep a scam-check prompt saved so you can paste any suspicious message and get a calm risk assessment.
• Never include your own OTP, password, or full card number, even when checking a scam; the signals are in the wording.
• Have AI inspect links without clicking, to catch look-alike domains like paypa1.com or names buried in fake addresses.
• Use a search-first tool like Perplexity to verify a company's legitimacy with cited sources before trusting it.
• AI is a strong second opinion, not proof; when in doubt, do not pay, do not share secrets, and verify independently.