AI keeps getting smarter. One day it was writing quirky email subject lines. And now it’s generating artwork, suggesting recipes, helping plan holidays, and even filtering your photos before you’ve taken them. It’s clever. It’s convenient. And honestly, pretty fun. There’s one thing AI always wants in return though – your data. 

With Safer Internet Day 2026 landing on Tuesday, 10 February, this year’s theme ‘Smart tech, safe choices – Exploring the safe and responsible use of AI’ couldn’t be more relevant. But it’s not about ditching your favourite AI apps and turning your phone into a brick. It’s more about using AI smartly, safely and without accidentally telling the internet your entire life story. And we’re here to show you how…

When an AI app is free, the developers still have costs. They’ve got to pay for everything from the servers and updates to the engineers and fancy tech working away behind the scenes. So, most AI tools pay these bills by learning from the things you upload or type. That means your selfie, prompt, and that question about how to fix your Wi-Fi could all be used to train the model. 

Now, to be clear, this doesn’t mean someone is manually scrolling through your chats like some sort of social media feed. Instead, it’s all automated. But your content still becomes part of the system’s ‘brain’, helping it to learn how humans write, speak and solve problems. 

Most major AI services (like ChatGPT and Gemini) now have an option tucked away in the settings. It’ll be called something like ‘Do not use my data for training’. Turn it on. It takes 10 seconds and makes a huge difference to your privacy. 

AI chatbots are getting really good at sounding like a human. Too good in fact. This makes it easy to start chatting like you would with a close mate or therapist (oversharing included). But AI chats really aren’t the place to start spilling your deepest secrets. 

You need to think about the ‘Dinner Party Rule’. If you wouldn’t stand up at a crowded table and announce it to everyone there, don’t type it into a chatbot. That includes your home address, work passwords, medical details, and bank login details. Pretty much anything you’d rather not see plastered on a giant billboard somewhere. 

And if you thought big companies tend not to make huge mistakes, think again. Samsung employees accidentally fed confidential source code into an AI tool, which then stored it for anyone to use. It’s proof that even the smartest people at massive brands can easily slip up. 

So, treat your chatbot like an enthusiastic stranger with excellent grammar. Helpful? Absolutely. Trustworthy enough for private details? Not so much.

Here’s where things get a little Black Mirror. AI is now smart enough to clone a voice from as little as three seconds of audio. Yes, seriously. And guess where scammers find those three seconds? You’ve guessed it – social media videos, podcasts, and TikToks. They’re a treasure trove of material for them to use and abuse. 

Suddenly, your mum might get a call from ‘you’, sounding panicked, saying you’ve lost your phone and need some money to get home right now. It’s believable. Too believable. And you can easily see why someone could fall for it. 

The fix is to set up a ‘Family Safe Word’ or phrase. It should be something weird and memorable. Or just plain impossible to guess. Think ‘parsnip’, ‘moon llama’, or ‘grandma’s slippers’. If someone rings you claiming to be a family member in trouble, ask them for the safe word. If they don’t know it, hang up and contact them another way. It’s straightforward but effective and can easily stop a scammer in their tracks. 

AI features might rely on your camera, microphone, contacts, or location. Or sometimes all four! Fair enough, sometimes they need that access to work properly. But other times, the apps can get a little, well, nosy. 

That’s where your phone’s Privacy Dashboard comes in. On both Android and iOS, you can check which apps have used your mic, camera, or location in the last 24 hours. If your calculator app suddenly decided to access your microphone at 2am, that’s your cue to hit ‘Revoke Permission’ immediately. 

This checks out pretty well for AI-powered apps, which sometimes request access for optional features you may never even use. Basically, if you’re not using the feature, you don’t need to share the data. Simple. 

It’s true that AI is fun, powerful, and super helpful. Saying that though, it needs boundaries. A few quick habits can keep your digital life safe without dialling down the enjoyment…

• Switch on the ‘Don’t train on my data’ setting
• Keep secrets and sensitive info out of AI chats 
• Set a family ‘Safe Word’ for any emergencies 
• Check your phone’s ‘Privacy Dashboard’ regularly

Does your family have a ‘Safe Word’ yet? If not, Safer Internet Day is the perfect time to pick one! Let us know in the comments if you’re getting smart with your tech this year.