01
Protect
Strip private, regulated, and confidential details first.
Responsible AI training
Safe AI use is a skill people can practice.
Responsible AI training should not be a policy slide everyone forgets. It should become a habit: protect the data, question the answer, keep a person accountable.
Safe AI decision habit
Make the review path visible before work leaves the screen.
Policy becomes practice.
02
Prompt
Give context, limits, and the decision standard.
03
Review
Check facts, assumptions, bias, and missing context.
04
Escalate
Move high-impact or uncertain work to a person.
Before someone uses AI
What data am I about to expose?
Who is accountable for the final answer?
What must be verified before this is used?
Risk and response map
Sensitive data
Do not paste private, regulated, customer, employee, or confidential information into an AI tool without authorization.
False confidence
Treat fluent output as a draft or suggestion until it is checked against context, facts, and policy.
Unclear authority
Know when AI can assist and when a person must decide, approve, or escalate.
Unreviewed reuse
Do not copy AI output into real work before checking source quality, assumptions, and missing context.
Governance habits
AI ethics training becomes useful when it changes everyday choices.
01
Name the data boundary before prompting.
02
Ask for uncertainty, assumptions, and review points.
03
Keep human approval visible on decisions that matter.
04
Document what was checked before output was used.