Scoring the Score: An Empathy, Trust and Accountability Lens for Auditing AI-Assisted Hiring Decisions

Hiring is now frequently mediated by algorithms that read resumes, grade recorded interviews, and rank applicants. Such systems are commonly evaluated on a single metric- how accurately they forecast performance. This paper argues that accuracy is a necessary but incomplete test, because it says nothing about how the resulting score is experienced by the applicant, understood by the recruiter, or defended before a regulator. We propose a three-part evaluation lens- Empathy, Trust and Accountability (ETA)- that assesses the decision the machine produces rather than only the model that produced it. Empathy concerns the dignity of the candidate's experience; Trust concerns whether the score is explainable and job-relevant; Accountability concerns whether it is bias-audited, lawful and open to human override under regimes such as the EU AI Act, NYC Local Law 144 and India's DPDP Act 2023. We express the lens as a small structural model with eight propositions, embody it in a free, open-source evaluation module, and illustrate it on a video-interview scenario. As future work, a survey-and-interview study is specified to test the propositions with accuracy held constant. The result is a conceptual framework and open-source artifact- an audit-ready instrument for responsible hiring AI.