How to avoid AI scams and misinformation online

AI has made two things significantly worse: scams and misinformation. Both are worth understanding.

AI-powered scams

The phone call from your ‘grandchild’ in trouble. The email from your ‘bank’ that sounds exactly right. The voice message from someone you know asking for help.

AI can now clone voices from a few seconds of audio. It can write phishing emails without the spelling mistakes that used to make them easy to spot.

The defence is the same as always: verify through a different channel.

If you get an unexpected call from a family member in trouble — hang up and call them back on a number you already have. If an email from your bank asks you to click a link — go directly to your bank’s website instead.

AI makes the scams more convincing. It doesn’t change the underlying logic of how to spot them.

AI-generated misinformation

AI can generate plausible-sounding false information at scale. Be sceptical of sources you don’t recognise. Check important claims against sources you trust. Use Perplexity when you need sourced, verifiable information.

When AI tools themselves are wrong

ChatGPT, Claude, Gemini — they all get things wrong sometimes. Confidently.

Treat AI output the way you’d treat advice from a very knowledgeable colleague: useful, worth taking seriously, but worth verifying before you act on it.

The simple rule

If something is unexpected, urgent, and asks you to act quickly — be suspicious. Take a breath. Verify through a channel you control.

— Anna