XIZOA / AI safety
Safety scorecard
A practical checklist for assessing whether an AI tool or workflow is appropriate for a particular task.
Core questions
Data: What information is collected, where is it processed, and how long is it retained?
Reliability: Can important outputs be checked against trustworthy sources or human review?
Control: Can a user export, delete, correct, or stop processing their information?
Misuse: Could the workflow enable privacy, security, discrimination, or other foreseeable harms?
Transparency: Are limitations, model behavior, pricing, and changes communicated clearly?
Use this scorecard as a screening tool, not as a certification or legal assessment.