How to Use AI to Fight AI Scams
The basic idea behind “using AI to fight AI scams” is sound: you can use an AI assistant as a second set of eyes on suspicious emails, texts, websites, and messages. AI can identify patterns such as urgency, impersonation, suspicious requests, unusual wording, and potentially deceptive links.
How I recommend using AI against scams
- Don’t click first—analyze first.
If you receive a questionable email, copy the text into ChatGPT and ask:
“Is this likely to be a scam? Identify every red flag and tell me what I should verify.”
You can also paste the sender’s address and the visible link, without clicking it.
- Ask AI to look beyond spelling mistakes.
This is important now. Good grammar is no longer evidence that an email is legitimate. AI can produce polished phishing messages, including personalized ones. - Have AI explain what the scammer is trying to accomplish.
For example:
“What information or action is this message trying to get from me?”
That can reveal that an innocent-looking “invitation” is actually trying to get you to log in, provide a phone number, install software, or make a payment.
- Verify independently.
This is the most important step. AI should help you investigate, not be the final authority. If an email supposedly comes from your bank, Amazon, Microsoft, a friend, etc., go directly to the organization’s known website or use a phone number you already have—not a link or number supplied by the suspicious message.
One important warning
Don’t paste sensitive information into an AI tool—especially passwords, Social Security numbers, credit-card numbers, account recovery codes, or private financial information.
And don’t rely on an AI detector to tell you whether a voice, photograph, or video is genuine. Current detection technology isn’t reliable enough to make that your only defense. Independent verification remains much safer.