Use AI with clear data boundaries
Choose suitable information to share and check outcomes before acting.
In plain language
Before using an AI service, consider what information the task needs and whether you may share it. Personal, confidential or restricted data requires an approved tool and an appropriate basis for use. Remove unnecessary details rather than assuming every assistant has the same privacy behavior.
Responsibility also includes checking accuracy, considering unfair outcomes and keeping a person accountable for important decisions. Describe uncertainty and limitations. A model can help draft a decision or compare options, while the person or organization using it remains responsible for the result.
To improve a customer email template, use a fictional customer and invented account details. You can test the wording without uploading a real customer record.
Try it yourself
Rewrite a request that contains a customer name, account number and address using fictional placeholders.
Go deeper · for developers and architects
Assess retention, access, data minimization and evaluation across affected groups. Permission to read a record does not automatically mean permission to send it to an external model.
Check your understanding
2 related questions, with an explanation for each answer.
Explore the original sources
Microsoft responsible AI curriculum ↗Report an issue
Tell the editors what needs correcting. Reports are reviewed and do not automatically remove content. Do not include personal or confidential information.
Finished this lesson?
Completion records your choice, not a measured skill level.
