Privacy in OpenAI ChatGPT

Claude (byAnthropic) prioritizes user privacy through encryption, limited employee access, and "Constitutional AI" training to prevent data leaks. While chats are private by default, Free/Pro/Max plans may use data to train future models if not opted out. Commercial users (API/Enterprise) generally retain stricter, non-training data policies.

I will show you how to turn off location and improve Claude in the settings of your ChatGPT account. You can follow along in the video or through step-by-step instructions.

Image generated by Midjourney ☝️

Watch the video here.

1. Turn OFF “Help Improve Claude” (training on your chats)

This stops your chats from being used to train Claude’s future models.

Steps (Web or Desktop)

1. Open Claude and log in.

2. Click your profile name or icon (usually bottom-left).

3. Click Settings.

4. Click Privacy.

5. Find “Help Improve Claude.”

6. Toggle the switch OFF.

✔ When OFF, your future chats and coding sessions won’t be used for model training.

⚠️ Note: Data that was already used in past training can’t be removed.

2. Turn OFF Location Metadata

Claude can use approximate location info (IP-based) for certain features. You can disable optional location metadata in settings.

Steps

1. Open Claude.

2. Click your profile icon/name.

3. Select Settings.

4. Go to Privacy (or Data & Privacy).

5. Find Location Metadata or Allow Claude to use approximate location.

6. Toggle the switch OFF.

✔ This prevents Claude from using your location metadata for contextual features.

⚠️ Important:
Some basic IP location info may still be used for compliance or regional requirements and cannot be fully disabled.

Michael Carmine
Founder & CEO | AIEducationalSolutions.org