Responsible use of AI tools
Learn a practical workflow for prompting, verification, attribution, and deciding when AI output is unsuitable.
Use AI for suitable tasks#
Generative AI can help brainstorm, rephrase, summarize provided material, create practice examples, or suggest questions. It is less suitable as the sole authority for high-stakes facts, private decisions, or work requiring guaranteed originality.
Protect information#
Do not enter confidential, personal, licensed, or organization-restricted data unless you have permission and the tool's handling is suitable. Remove unnecessary identifiers and follow the relevant policy.
Verify output#
Fluent language can hide errors. Check calculations, names, dates, quotations, and citations against authoritative sources. If a cited source cannot be found or does not support the claim, do not use it.
Preserve learning#
Use AI to explain reasoning or provide feedback, then solve and restate the idea yourself. Submitting generated work without understanding may violate rules and prevents learning.
Bias and representation#
Outputs can reflect gaps or stereotypes in data and prompting. Review who is represented, which assumptions are made, and whether alternatives were ignored.
A responsible workflow#
Define the task, limit sensitive input, request a useful format, verify important details, revise with your own judgment, disclose assistance where required, and remain accountable for the final result.
Key points
- AI output can be fluent and still be inaccurate or fabricated.
- Do not submit confidential or personal data without authorization and suitable safeguards.
- Verify important claims with authoritative sources.
- The human user remains responsible for decisions and submitted work.