AI Hiring Filters Reward Applicants the Model Already Recognizes

The institutional risk is that hiring systems can convert prior online exposure into evidence of fit. The Bluesky question about models favoring applicants they already know ^[1] and the documented pattern of model-written resumes clearing screens more often through the same model used by employers point to the same machinery: the filter rewards forms of expression already familiar to it. The loser is not only the underrepresented applicant; it is the employer that mistakes machine familiarity for job relevance.