Is ChatGPT safe to use?
General-purpose AI assistant from OpenAI for conversation, writing, coding, image generation, and more. On an Apple-silicon Mac, the ChatGPT desktop app can optionally install an Apple Messages plugin. In Codex or ChatGPT Work, that plugin can read and search iMessage, SMS, and RCS chats and send messages through the Messages app; it does not work in a regular ChatGPT chat or on mobile.
Last reviewed: 12 September 2026
The short answer
Mostly, with care.
- We rate ChatGPT Use With Caution (privacy risk: Medium).
- Trains on what you type: Yes, unless you switch it off — and you can. How to switch it off
Using a few AI tools at home? All the family AI privacy settings in one place, ready to print.
Does ChatGPT train on what you type?
Yes, unless you switch it off — and you can.
Can you turn ChatGPT's training off?
Model training: open Settings > Data Controls and turn off Improve the model for everyone. That setting also covers Codex tasks on a personal plan. Use Temporary Chat when you do not want a conversation in history. Apple Messages: if you did not install the plugin, it has no Messages access. If you installed it and no longer want it, uninstall the Apple Messages plugin and review ChatGPT’s permissions under macOS System Settings > Privacy & Security. If you keep it, use Allow once. To remove persistent send approval, open ChatGPT Settings > Computer use, select Manage next to Messages, then remove the chat under Always allowed to send. Managed-workspace administrators can disable Apple Messages through the Computer Use control.
Is what you put into ChatGPT private?
OpenAI collects the prompts and other content you provide, its responses, account information, device information, and usage data. For personal ChatGPT accounts, content may be used to improve models unless you turn off Improve the model for everyone. Turning it off keeps chats in history but stops new conversations and personal-plan Codex tasks from being used for training. Temporary Chats are not used for training, do not appear in history, do not create memories, and are deleted from OpenAI’s systems after 30 days, though they may be reviewed for abuse. OpenAI says Business, Enterprise, Edu, and API data are not used for training by default. The Apple Messages plugin is optional and Mac-only. OpenAI’s current plugin documentation says it can read and search chats on the Mac and send messages as you after you grant permissions. Per-send approval is the default. Choosing Always allow sending to this chat removes the approval step for later messages to that thread. OpenAI warns that persistent approval removes your final chance to review a message before ChatGPT sends it as you. OpenAI told TechCrunch that the plugin does not build a full index of the inbox, runs locally, and reads messages after a specific request; it also said message content is stored locally unless the user saves the conversation to the cloud, when ordinary cloud retention rules apply.
So, is ChatGPT safe to use?
Mostly, with care. We currently rate ChatGPT "Use With Caution" (privacy risk: Medium). Worth reading the points above before you share anything personal.
Our take
“I heard ‘it runs locally’ and nearly translated that as ‘it cannot send a family text as me.’ Those are different sentences. I am keeping the approval step.”
The official documents behind this review
We read these so you don’t have to. Check them yourself any time.
Wondering how we decide these ratings? Read how we review AI privacy.
Keep reading
From the Unspun Blog
- The twenty minutes ChatGPT doesn't countThe five-minute AI draft can hide fifteen minutes of checking. A survey of 6,000 workers gave that unpaid work a name: botsitting.
- ChatGPT for Financial Services is for banks, not your tax folderOpenAI launched ChatGPT for Financial Services for banks. It is not a household money app, and it is not a reason to paste your bank statements.
- You shouldn't have to learn Spark vs Chat vs Cowork to send an emailAI apps keep asking ordinary people to choose between Spark, Daily Brief, Chat, Work, and Cowork. The app should do that homework.
Knowing what’s safe is half of it.
The other half is knowing what to do with these tools. We teach that part: tutorials, prompts, plain English, no judgment. $7 a month, and the first seven days are free.
Show Me How This Works →Start free. Cancel anytime. No judgment, ever.