AI Safety for Beginners: What Not to Share (and Why It Matters)

Part 4 (Final) of the AI Series for Midlife Adults

Over the past few weeks we’ve covered a lot of ground together: why midlife adults should learn AI, why it isn’t making you dumber, and how to make sure it doesn’t replace your unique voice. For this final post, we’re talking about something just as important: how to use AI safely, so you can enjoy all the benefits without putting yourself at risk.

Before you feel any hesitation, this isn’t meant to scare you away from exploring AI. Think of it like any other online activity. There are some basic common-sense practices that keep you protected, and once you know them, they become second nature. Let’s walk through them together.

AI is a Tool, Not a Vault

There are dozens of AI platforms to explore, ChatGPT being the most well-known.  Each with specialized uses like marketing, writing, coding, medical information, or even life coaching. These are genuinely useful tools, as long as you understand one important reality: anything you type into an AI chat window leaves your device.

Unlike a private journal or a locked filing cabinet, AI platforms are cloud-based services. When you type something in, that information is sent to and stored on external servers. Depending on the platform’s privacy settings, your data could potentially be used to train future AI models (some platforms do this by default unless you opt out), stored on servers that could theoretically be breached, or in some cases reviewed by company staff for safety and quality purposes.

This isn’t about assuming bad intent from AI companies. It’s simply the reality of how cloud-based technology works, and the same logic that tells us not to share our social security number over email applies here too.

What You Should Never Type Into an AI Platform

Here’s a practical rule of thumb: if you wouldn’t write it on a sticky note and hand it to a stranger, don’t type it into a chat window.

Specifically, avoid sharing:

• Your full name combined with your address, birthdate, or phone number
• Social security numbers, tax ID numbers, or passport details
• Financial account numbers, credit card details, or banking information
• Passwords or answers to security questions
• Medical records or health documents with your personal identifiers on them

The good news is you rarely need this specific information to get value from AI. Use generalities instead.  For example:

• If you’re asking about lawn care, you can mention your state and grass type without giving your exact address.
• If you have a health question, you can type in the specific result or symptom like “my A1C was 6.0, what does this mean for my metabolic health?”, without uploading a document that has your name, date of birth, and doctor’s details on it.
• If you want help creating a budget, type in the categories and amounts rather than uploading a bank or financial statement.

A little creative thinking goes a long way here, and you’ll still get excellent results.

What About Photos?

This is a legitimate concern and slightly more nuanced than text. When you upload a photo to an AI tool, a few things can happen depending on the platform. The image may be stored and potentially used in training future AI models. Facial recognition data can be extracted from photos of faces, even unintentionally. And in some cases, particularly with lesser-known apps, photos have been used to generate synthetic images without the user’s knowledge or consent.

The bigger risk usually isn’t the major platforms like ChatGPT or Google, it’s the many third-party apps built on top of AI technology that have much looser privacy policies and less accountability. Before uploading any photo to an AI app, especially a newer or less familiar one, it’s worth taking two minutes to skim the privacy policy and look for language about how images are stored and used.

As a general rule: avoid uploading photos that contain other people without their knowledge, photos with location data embedded (most smartphone photos carry this), and anything you would consider sensitive or private.

Simple Habits That Protect You

Everyone has to decide their own comfort level with AI, but these baseline habits are worth making routine regardless of which platforms you use.

Check your privacy settings. Using ChatGPT as an example: go to your profile name at the bottom left, click Settings, then Data Controls. Under “Improve the model for everyone,” toggle this OFF. This opts you out of having your conversations used for training. Most major platforms have a similar setting, so make this the first thing you do when you set up any new AI account.

Use a separate email address for AI tools. This is an easy extra layer of protection. If that account is ever involved in a data breach, your primary email and everything connected to it, stays separate.

Opt out of data sharing wherever possible. Beyond the training toggle, check each platform for settings related to location data, usage analytics, and third-party sharing. When in doubt, turn it off.

Be mindful of app connections. Many AI platforms allow you to connect third-party apps,  things like your calendar, email, or even financial tools. Be selective. Avoid connecting AI platforms to anything that contains sensitive personal or financial information unless you’ve thoroughly reviewed the security practices involved.

Keep your AI apps updated. Security patches are released regularly, and running an outdated version of any app leaves you more vulnerable. This applies to AI tools just like any other software on your phone or computer.

A Resource I Trust for Staying Safe Online

For me, staying safe online is about understanding the risks and making informed choices, not living in fear of technology. I genuinely believe AI is a valuable tool that can make everyday life easier and more productive, especially for those of us running online businesses.

That said, cybersecurity is its own area of expertise, and I’m not going to pretend to be an expert in it. For that, I turn to people who actually are. One person I personally follow and recommend is Caitlin Sarian, known as @cybersecuritygirl on Instagram. She breaks down data breaches, online safety tips, and the latest in cybersecurity in a way that’s accessible and not overwhelming, which is exactly what most of us need. I highly recommend giving her a follow if you want to stay informed from someone who truly knows this space.

Wrapping Up the Series

And that brings us to the end of our AI series. We started with why learning AI matters for midlife adults, moved into why it isn’t eroding your intelligence, talked about protecting your voice and thinking, and now we’re closing with the safety practices that make it all sustainable.

AI isn’t something to fear, but like any powerful tool, a little awareness makes all the difference. You now have enough to get started confidently, stay protected, and keep learning at your own pace.

Interested in learning how Digital Midlife uses AI on a daily basis? Drop your questions in the comments or contact me directly, I’d love to know if a follow-up post on my personal AI workflow would be helpful.

Related Posts:

• Part 1: Why Midlife Adults Should Learn AI
• Part 2: AI Isn’t Making You Dumber
• Part 3: How to Stop AI From Replacing Your Voice
• How I Created My AI Avatar