Lesson 1 of 6 · 9 min
How AI tools work
What happens to what you type into an AI tool, who can see it, and three questions to ask before you use one.
Your prompt goes somewhere
When you type into an AI chatbot, or an AI feature inside another app, your text is usually sent over the internet to a company that runs the AI model. Depending on the product and its settings, that company may store your text, let staff review it to improve the service, or use it to train future models.
So treat anything you paste into a prompt as shared with that company. If you paste a person's name with their medical history, their case notes or their grades, you have disclosed that information to whoever runs the model.
The model predicts, it does not look things up
AI models write by predicting likely next words, based on patterns in the text they were trained on. That is why they are fluent and fast, and also why they can state something false in a perfectly confident tone. Lesson 2 comes back to this.
Personal accounts versus approved tools
A free personal account and a work-approved version of the same tool can have very different rules. Business and education versions often come with a contract that says your data is not used for training, is kept for a limited time and is protected. Free consumer accounts often have none of that.
The tool's name is not what matters. What matters is whether your organization has an agreement covering it.
Three questions before you use any AI tool with personal information
If the answer to any of these is no, or I do not know, keep personal information out of it.
- Has my organization approved this tool for this kind of information?
- Is there an agreement that says the data will not be used to train models?
- Do I know how long my prompts are kept, and who can see them?
Lesson quiz
Get 3 of 4 right to pass.