Can AI automatically reject candidates?
Not with Talent Pronto. Anna scores every interview against the rubric your team defined and ranks candidates into a shortlist, but she never rejects anyone. No candidate is auto-rejected by the system — your team reviews scores, reasoning, and transcripts, and makes every decision about who advances.
Why this question matters
Auto-rejection is the failure mode that gives AI hiring tools a bad name: a system silently filters out candidates based on resume keywords or opaque scores, no human ever sees them, and nobody can explain why. That design creates real legal exposure — regulations like NYC Local Law 144 and Illinois' AI Video Interview Act exist largely because of it — and it costs employers strong candidates whose value doesn't fit a keyword filter.
So the question isn't whether AI can auto-reject candidates. Some systems do. The question is whether your screening tool is designed so that it can't.
How Talent Pronto separates scoring from deciding
Anna evaluates; your team decides. The division of labor is strict:
- Anna interviews every applicant in a structured, conversational text interview — the same questions, in the same way, for every candidate in a role.
- Every answer is scored against your rubric. Knockout questions (work authorization, required certifications, availability) are pass/fail. Behavioral questions are scored 0–100 with written reasoning you can read and challenge.
- Candidates are ranked, not filtered. Every applicant appears in your dashboard with their score, interview highlights, and full transcript — including the ones who scored poorly and the ones who failed knockouts. Nothing is hidden and no one is removed.
- Your team makes every call. Advancing, rejecting, and hiring are human decisions, made with Anna's evidence in front of you.
Why explainability is the other half of the answer
A "no auto-rejection" policy only means something if humans can actually review the AI's work. Every Talent Pronto interview produces a complete audit trail: the questions asked, the candidate's answers, per-criterion scores, and the written reasoning behind each score. Scoring is EEOC-aligned and identical for every candidate in a role. If a score looks wrong, you can read exactly why it was given and overrule it. See security, compliance, and fair hiring for the full picture.
Common follow-up questions
What happens to candidates who fail knockout questions?
They're marked as not meeting the requirement, with the specific question and answer recorded — but they stay visible in your pipeline. Your team can review them and still choose to advance anyone.
Can I configure Talent Pronto to auto-reject anyway?
No. Rejection is always a human action. That's a deliberate design constraint, not a missing feature.
Does a ranked shortlist just move the bias into the scores?
The rubric is yours: your criteria, your weights, applied identically to every candidate. That's more consistent — and far more auditable — than unstructured phone screens, where interview bias typically lives. Every score comes with reasoning your team can inspect.
