Verify on a Second Channel

This is the single most effective habit against essentially every AI-enabled scam on this page, because it works regardless of how convincing the voice, video, or message is. The principle is simple: if someone contacts you with an urgent request — especially involving money, credentials, or sensitive information — you independently verify it through a channel you already trust, rather than one they just gave you.

In practice, that means hanging up and calling a number you already had saved, navigating to a website by typing the address yourself instead of clicking a link, or messaging a coworker through your company's normal system rather than replying to the email that supposedly came from them.

The reason this works even against very convincing fakes is that it sidesteps the whole question of "does this look real?" entirely — you're not trying to spot a fake, you're just refusing to let the contact itself be the source of truth for how to verify it.

• The rule: Never verify using contact information the request itself provided.
• In practice: Hang up and call back, or navigate to the site directly — don't click through.
• Works against: Voice cloning, deepfakes, phishing, fake support — nearly everything on this page.
• Make it normal: Tell your family or team this is expected, not rude or paranoid.