Review the tools
Download the four practitioner artifacts that anchor the assessment.
Open the library →Doctoral research · Responsible AI
BEYOND
BIAS
The Protective Alignment Framework
Engineering gender fairness in AI-driven credit decisioning—from data design to accountable deployment.
Confidential · Voluntary · Approximately 15–20 minutes
Why this research matters
Bias persists in the handoffs.
Historical credit data carries unequal access, discriminatory lending patterns, pay gaps, and employment inequities. Removing sex from a model does not remove the effect: address, occupation, income stability, transaction activity, and device use can preserve those patterns as proxies.
Organizations often have capable data science, compliance, risk, and executive teams. Yet fairness work remains fragmented across them. PAL treats detection as a trigger—not a destination—and governs the sequence from correction through monitoring, explanation, and evidence.
Your professional judgment will test whether that sequence and its tools are relevant, feasible, and defensible in practice.
Share your judgment
Rate their relevance and feasibility from your professional experience.
Begin the survey →Shape the framework
Your confidential response informs the framework, dissertation, and practitioner book.
About the research →