Lesson 6: Staying safe: privacy and AI scams
Treat a chatbot like a postcard, not a sealed letter.
🧸 The simple picture
When you mail a postcard, anyone who handles it along the way could read it. So you'd never write your bank password on a postcard.
Typing into an AI chatbot is a bit like that. Depending on the app and your settings, what you type may be stored and, in some cases, used to improve the AI. So only write things you'd be okay with a stranger seeing.
🔑 Key ideas
- Never share: passwords, full Social Security numbers, bank or card numbers, medical record numbers, or other people's private info.
- Check the app's settings. Many chatbots let you turn off chat history or opt out of your chats being used for training.
- Voice cloning scams: AI can copy someone's voice from a short clip. If "a family member" calls in a panic asking for money, hang up and call them back on their real number. Consider a family code word.
- Deepfakes: AI can create fake but realistic photos and videos. If something seems shocking, check whether trusted news sources are reporting it before you share.
- Scam emails and texts are getting better-written thanks to AI. Bad spelling is no longer a reliable warning sign. Be suspicious of urgency ("act now!") and requests for gift cards, wire transfers, or crypto.
🛠️ Try it (2 minutes)
Two quick safety wins today:
1) Open the settings of the chatbot you use most and find the privacy or data controls. Decide whether you want chat history or training use turned on.
2) Text your family: "If anyone ever calls pretending to be me in an emergency, ask for our code word." Pick a word together.
📝 Quick quiz
Tap an answer to check it.
1. Typing into a chatbot is most like…
2. You get a panicked call from a "grandchild" asking for money. Best move?
3. Which is a big warning sign of a scam?