The Eligibility Checklist That Quietly Excluded People

In my federal grants compliance days, I worked with database-driven eligibility checks — rules that decided who qualified for help. The rules looked neutral. Every applicant answered the same questions and the system applied the same logic.

But neutral rules can still screen people out unevenly. A documentation requirement that most applicants could meet easily was nearly impossible for one group — not because the rule mentioned them, but because their circumstances didn't fit the form. Nobody designed it to exclude. The design just never considered them.

In plain English: a rule can be fair on paper and unfair in practice. AI eligibility and screening tools do exactly the same thing, at a much bigger scale.

The lesson I carry from that work: don't just check whether the rule is neutral. Check who it actually lets through. Run the outcomes by group, the way I learned to reconcile grant reports — line by line, no assumptions.

Fairness lives in the results, not in the intentions. The checklist doesn't know who it excluded. That's your job to find out.

I explain frameworks; I don't give legal counsel.