Misinformation, deepfakes, and 'AI tricks' (how to stay safe)

AI can be used for good (helpful writing, learning, accessibility). But it can also be used to create convincing misinformation. This page gives you simple habits to protect yourself and your family.

What is misinformation?

Misinformation is false or misleading information. Sometimes it is shared by accident. Sometimes it is shared on purpose.

How AI changes the game

AI makes it easier to create content that looks or sounds believable, including:

• fake news articles or social media posts
• fake emails that look professional
• fake images (including 'photos' of events that never happened)
• fake audio (a voice that sounds like a real person)
• fake video (deepfakes)

Deepfakes explained simply

A deepfake is media (usually video or audio) that has been generated or altered using AI to make it look like someone said or did something they did not.

Common deepfake risks for everyday people include:

• A voice message that sounds like a family member asking for urgent money
• A fake video of a public figure claiming something shocking
• A fake 'news clip' shared on social media

The golden habit: slow down

Scams rely on urgency. The best protection is slowing down before you click, pay, or reply.

A simple verification checklist (print this)

1. Pause. Do not click links or open attachments yet.
2. Look for the pressure tactic: urgency, threats, or guilt.
3. Check the source: is it from the official channel you normally use?
4. Verify independently: call the organisation using a number from a bill, card, or official website (not the message).
5. Ask a second person for a sanity check if you feel unsure.

Family safety: the 'code word' trick

A simple family safety trick is to agree on a private code word or phrase for genuine emergencies.

• If someone messages or calls asking for urgent money, ask for the code word.
• If they cannot answer, assume it may be a scam and verify through another channel.

Using AI to protect yourself (yes, you can use AI against scams)

AI can help you spot red flags, but you must use it safely. The key is redaction: remove personal details before pasting a message.

I received this message (personal details removed):

[PASTE MESSAGE]

1) List scam red flags.
2) Suggest safe next steps.
3) Tell me what NOT to do.
4) If the message mentions a company, tell me how to verify using official contact details.

Common 'too good to be true' patterns

• Unexpected refunds or prizes
• Threats about accounts being closed
• Requests to move the conversation to a different app quickly
• Requests for gift cards, crypto, or unusual payment methods
• Links that look similar but not identical to real websites

How to check photos and videos (practical tips)

You do not need specialised tools for most cases. Start with these habits:

1. Check where it came from. A random post is not evidence.
2. Look for odd details: strange hands, blurry text, inconsistent shadows, 'wobbly' edges.
3. Check if other trusted sources are reporting the same thing.
4. If it matters, wait. Real news spreads across multiple reputable outlets.

What to do if you already clicked or responded

Do not panic. Take calm steps.

• If you entered a password: change it immediately and enable two-factor authentication if available.
• If you shared banking details: contact your bank using official contact details.
• If you installed an app: uninstall it and run a security scan if available.
• If money was sent: contact your bank quickly. Time matters.

A short reality check about AI 'facts'

Even when there is no scam, AI can still be wrong about facts. For important information:

• Ask it to list assumptions and uncertainty.
• Ask for a verification checklist.
• Confirm critical details on official sources.

Screenshots