Responsible Use Model
RESPECT
Seven principles for responsible AI. Use them to guide the people, decisions and practices behind AI-assisted work.
Seven principles
Responsible
Know your own responsibilities and how to report an error or concern.
Equitable access
Ask for help or an agreed alternative if a tool or process creates a barrier.
Safety
Verify the parts that matter to this task. Raise uncertain or harmful results before using them.
Privacy
Check the account, data and sharing rules before entering information or connecting a source.
Efficient and effective
Count the time spent checking and correcting. Tell the owner when the process makes work harder.
Control
Keep work within the agreed limits. Contact the owner if you cannot correct or pause it yourself.
Transparency
Use the agreed disclosure and make the route to human help clear.
Apply all seven together. People need a responsible owner, appropriate support and a clear way to question or correct the work.
Apply it to one task.
Choose a recurring task and involve the people doing and receiving the work. Answer each question with a decision, evidence or an unresolved issue.
Responsible
Who owns this task, approves its AI use and responds when someone reports a problem?
Equitable access
Who will use this process or be affected by it, and what access, training or accessibility support do they need?
Safety
Which errors could harm people or the organization, and what checks or stop conditions will address them?
Privacy
Which tools, accounts and information are allowed, and where may inputs and outputs be stored or shared?
Efficient and effective
How will we compare quality and total effort with the current process, including review and correction?
Control
Which actions may run automatically, which need approval and who can pause, correct or override the system?
Transparency
Who needs to know that AI is involved, and how will we explain its role, limitations and the route to human help?
Turn the answers into working rules. Record the owner, approved tools and information, required checks, stop conditions and disclosure. Assign someone to resolve each gap before an authorized owner approves the process.
RUDI’s published framework: What Is Human-Centered AI? The RESPECT Framework