A generative AI system can often give a better response when it receives useful context.
That does not mean you should provide every piece of information available to you.
A responsible rule is:
Share only the information needed for the task and permitted for the tool you are using.
This idea is called data minimization.
Be careful with information that belongs to you or someone else.
Examples can include:
Whether a particular system is approved for a particular kind of information depends on the rules of your institution or workplace.
When in doubt, use the approved policy and tool guidance.
Suppose you want AI help rewriting a technical support message.
The original note says:
The student ID is probably not needed for a writing task.
A minimized version might say:
If identity is not needed for the task, remove it.
Suppose you need help understanding a spreadsheet formula.
Instead of uploading a real private employee list, you can often create a small example such as:
| Employee | Hours |
|---|---|
| Person A | 8 |
| Person B | 6 |
The formula question can be answered without exposing the original records.
Technical learning often works well with fictional or de-identified examples.
Different AI services can have different:
Do not assume that because one AI system is approved for a task, every AI system is approved for the same data.
Use the tools and practices authorized by the organization responsible for the information.
It can be easy to treat an AI chat like a personal scratchpad.
Instead, think before pasting.
Ask:
If the answer suggests the information is unnecessary or restricted, do not include it.
A screenshot can accidentally expose:
Before sharing an image with an AI tool, inspect the whole image.
Crop or remove information that is not needed and that you are permitted to remove.
Code may contain:
Do not paste code into an AI system simply because it is "only code."
Use the same data-minimization question:
You may find a tool helpful.
That does not automatically mean you are allowed to upload institutional, employer, or client information to it.
Follow the rules that govern the information.
Responsible AI use includes knowing when not to use AI for a particular input.
Before placing information into an AI system, ask:
If the task can be completed with less data, use less data.