TUESDAY TUTORIAL
One useful thing you can do with AI today.

Good morning.
I keep hearing from readers who assume their AI conversations are private by default. They usually aren't. This one is worth five minutes of your time.

First time reading? Get your own free subscription here.

Most people don't spend much time thinking about what happens after they hit send in an AI chat. Maybe they should.

A recent Census Bureau survey found that only 11% of Americans feel like they actually control how AI tools use their data. Barely one in five feel like they even know when they're using AI at all, given how many products build it in without asking first.

If that sounds uncomfortably familiar, there is something you can do about it.

Here is the good news: most of the major AI tools do give you a real off switch. It is just buried a few menus deep, and it works differently depending on which one you use. Below is where to find it, platform by platform, plus what these settings actually stop and what they do not.

One thing before you start clicking: these menus move without warning. If you're reading this a few months from now, expect the exact click path to shift a little. The setting itself should still be there.

ChatGPT

Open ChatGPT, click your profile icon, go to Settings, then Data Controls. Turn off "Improve the model for everyone." New conversations after that point will not be used to train future models.

If you want to talk about something sensitive right now without changing account settings, use Temporary Chat instead. It is never used for training and disappears within 30 days.

Google Gemini

Go to myactivity.google.com/product/gemini while signed in. Find "Keep Activity" and turn it off. This stops Google from storing your prompts long-term or using them to improve its AI.

One catch worth knowing: even with this off, Google holds your interactions for up to 72 hours to actually run the service. That window is normal and not something you can turn off.

Meta AI (Facebook and Instagram)

Open Settings and Privacy, then Privacy Center. Look for the section on Generative AI, then use Meta's objection request form to ask that your information not be used for AI training.

Worth knowing: this carries real legal weight if you live in the EU, UK, or a handful of other regions with strong privacy law. In the US, you can still submit the request, but Meta is not required to honor it.

Claude

Click your profile name, go to Settings, then Privacy. Find "Help improve Claude" and switch it off. Full instructions live on Anthropic's Privacy Center if a menu has moved by the time you read this. Incognito conversations are never used for training, with or without this toggle changed.

What these settings do not do

None of these toggles reach back in time. If a past conversation already went into a training run, turning off the setting today does not pull it back out. Deleted chats can also sit in backend systems for a while, sometimes 30 days, sometimes longer, for abuse monitoring or legal reasons.

And turning off training is not the same as turning off AI altogether. Gemini can still summarize your Gmail, Meta AI still shows up in your feed, and ChatGPT still remembers things in that conversation. You are only closing the door on future training, not the tool itself.

That distinction matters. It is the difference between a company using your grocery list to build a better assistant and a company simply having your grocery list.

This week: pick the one AI tool you use most. Spend five minutes finding its toggle. That is the whole task.

Try This With AI

What it does: Builds a personal checklist of every AI tool you actually use, so you always know where to go check.

Before you begin: Think of the AI tools you use at least once a month. ChatGPT, Gemini, Meta AI, Claude, or others.

Try it now: ChatGPT · Claude

Then ask this: "Turn this into something I can check every six months, since these settings sometimes reset after an update."

READER POLL

WHERE TO GO NEXT
More on this topic, from sources worth your time:

FROM OUR PARTNER

Most AI tools are a trade — your data for intelligence. Norton Neo breaks that deal. Powerful built-in AI, anti-fingerprinting, VPN, and ad blocking come standard. No setup. No add-ons. No compromises. Search, summarize, and write with AI that works inside your browser and stays there.

Browse Smarter, Stay Private