Human oversight
People retain final authority in high-impact hiring, performance, promotion, and compensation decisions.
AI GOVERNANCE
We manage AI with human oversight and trust principles
CADRO treats AI as a controlled decision-support tool, not an authority that independently decides matters involving employees, candidates, or customers.
Core principles
Data minimization
Personal data that is not necessary for the stated purpose is not introduced into an AI workflow.
Explainability
Inputs, rules, limitations, and review steps should be understandable where practicable.
Traceability
Use, review, correction, and override steps are recorded according to product requirements.
Risk management
AI use cases are classified by risk. High-impact use requires legal/privacy review, human approval, bias assessment, an appeal route, and stop conditions before live use.
AI lifecycle
- Govern: define owners, policy, access, and accountability.
- Map: document purpose, data flow, and affected people.
- Measure: evaluate quality, errors, bias, trust, and business outcomes.
- Manage: reduce risk, review incidents, and renew the use decision.