Keep private work private
Choose the right place before you upload
After this, you'll be able to decide whether a file belongs in normal chat, Temporary Chat, a Project, a GPT, or outside ChatGPT.
Before you start
Complete Set memory and custom instructions first.
The idea
Privacy is a placement decision before it is a prompt decision. The same file can be safer or riskier depending on where you put it and what your account settings allow.
Here is the before and after: before, ChatGPT is answering a loose request. After, you can decide whether a file belongs in normal chat, Temporary Chat, a Project, a GPT, or outside ChatGPT.
Now try it use the exercise prompt on one real task from today. Keep the answer small enough to check before you reuse it.
You are ready when the task, source, surface, and proof are all visible.
Try it (12 min)
Watch out for
- Uploading first and thinking about data controls afterward.
- Assuming Temporary Chat solves every company-policy issue.
- Putting sensitive reference files into a GPT because it feels organized.
- Letting memory absorb temporary private details.
Paste this into Claude
Help me decide how to handle this ChatGPT file or data task. Task: [what I want ChatGPT to do] Data involved: [file type and sensitivity] Account type: [personal, work, school, unknown] Need future reuse? [yes/no] Need memory to use this later? [yes/no] Need sharing with others? [yes/no] Give me a placement decision: - normal chat - Temporary Chat - Project - GPT knowledge - approved workspace path - do not upload Then give me the safer prompt or redaction plan.
What a good response looks like
Placement: do not upload the full file. Redact client names and paste only the decision table. Prompt: 'Using only this redacted table, find missing owners and deadlines. Do not infer company names or add external context.'
What good looks like
- The answer starts with placement, not prompt wording
- Sensitive data is either reduced, governed, or kept out
- Future reuse and memory are considered separately
- The final prompt limits what ChatGPT should use
When this breaks
- Breaks when convenience beats placement because sensitive context lands somewhere it can be reused later.
- Breaks when policy and account type are unknown because the same prompt may be acceptable in one workspace and wrong in another.
AI can help with this
Use ChatGPT to help you you can decide whether a file belongs in normal chat, Temporary Chat, a Project, a GPT, or outside ChatGPT. Start with the exercise prompt and your real input. Ask for one draft, then check it against this proof: The answer starts with placement, not prompt wording. Accept only the version you can verify yourself.
You can now
You can choose where data should go
- You can distinguish retention from memory
- You can reduce data before using ChatGPT
- You can stop when policy is unknown
Key takeaways
Private work starts with placement. Decide where the data belongs before you decide what prompt to write.
- Temporary Chat, normal chat, Projects, and GPTs have different reuse patterns.
- Memory can use saved context when enabled, so review it.
- Company work follows workspace policy before personal convenience.
- Redaction and local checklists often solve the task without uploading the full file.
Go deeper