Reference reading
Favoritism
Unjustifiably favoring certain people's claims, evidence or opportunities in AI-assisted selection and review.
The review is automated. The favored application already has its own shortcut.
Persistent, unjustified favoritism in the weight given to people's claims, evidence or opportunities.
Statistical disparity alone does not establish an inner habit; relevant distinctions and correction of disadvantage can be justified.
How it takes hold
The review is automated, but the favored application has a private route. Preference survives in thresholds or exceptions beneath uniformly processed output.
Favoritism persistently gives unjustified weight to certain people's claims or opportunities. Statistical disparity requires investigation rather than an automatic diagnosis of motive.
Those without favored standing can encounter a similar-looking process that weighs their circumstances less. Automation may make preference harder to contest.
Finding the distinction
Illustrative situation: similar applications enter review.
Too partial: a favored applicant receives a private shortcut.
Fitting: differences have relevant defensible reasons.
Overcorrection: an appropriate accommodation is prohibited.
Dishonesty can conceal preference; Favoritism concerns its unjustified weight. Negligent design can cause unfairness without this motive.
Relevant accommodations and correction of disadvantage can justify different treatment. Fairness is not identical processing regardless of circumstance.
Counterfeit of Fairness: Identical-looking processing hides a private exception for a favored application. Genuine Fairness examines the relevant weight and available routes, not only the surface rule.
A response that restores something
Review the reasons, admitted evidence and actual exceptions. Uniform appearance is insufficient.
Correction should reach both an unfair decision and a rule that would repeat it.
Fairness admits relevant differences while refusing private shortcuts that cannot be defended to those excluded.
Other perspectives
raises the value and function of each member of the community
Jane Addams · Democracy and Social Ethics
Addams makes inclusive regard a positive ethical aim rather than treating uniform machinery as sufficient fairness.
we can now examine our own conduct with his candour and impartiality
Adam Smith · The Theory of Moral Sentiments
Smith connects fair assessment to stepping outside our immediate interest, not to erasing relevant differences.
Read and compare
What relevant reason justifies this different weight?
Fairness · Integrity · Discernment
These are constructive virtue readings, not a list of automatic opposites. This reference article does not score character or assign a vice to its reader.