How employers use AI to score interviews comes down to three set-ups: an AI interviewer asks the questions and rates your answers, AI scores a one-way video you record alone, or an AI note-taker records your conversation with a human interviewer and drafts the notes and scorecard. In all three, the software mostly works from a transcript of what you said, compares it with criteria the employer chose for the role, and passes a score or summary to a person who decides what happens next.
The set-up you get changes what is scored, who reads the result and how much the AI's output matters. This guide walks through each one using what vendors publish about their own tools, checked in October 2026. For definitions of the formats themselves, start with what an AI interviewer is. One rule applies to every set-up below. Take the assessment yourself and follow the employer's rules — use CoPilot Interview to practice beforehand.
How employers use AI to score interviews: the three set-ups
| Set-up | Who asks the questions | What's scored | What it produces | Who sees it |
|---|---|---|---|---|
| AI interviewer (voice or video bot) | The AI, including follow-up questions | Each answer, against criteria built from the job description | A recording, a transcript, a rating per question or skill, and a summary | Recruiters, usually inside the employer's hiring system |
| AI-scored one-way video | Recorded or written prompts; you answer alone against a timer | Your transcribed answers, against a rating scale for each competency | Competency scores and an overall result compared with other applicants | Recruiters and hiring managers |
| AI note-taker plus a human scorecard | A person | The interviewer rates you; the AI organizes the evidence | A transcript, structured notes and a draft scorecard | The interviewer, who reviews and submits it, then the hiring team |
Some employers use more than one. For example, an employer might use a one-way video or an AI interviewer early in the process, then interviews with people, where a note-taker writes up what you said.
Set-up 1: an AI interviewer asks and rates
A conversational AI interviewer talks to you, listens and asks follow-ups. HireVue describes how its AI Interviewer works from the employer's side, as of October 2026 (HireVue AI Interviewer):
- The AI suggests questions and answer rubrics based on the job description, and the employer customizes and approves them.
- Qualification questions are scored as meets or does not meet. Skill questions use a custom 1-to-5 rubric or HireVue's own validated competencies, with scores tied to behaviorally anchored rating scales.
- The AI evaluates each response against the role's criteria, highlights whether it meets them and explains why.
- Recruiters see the scores in their applicant tracking system, and HireVue says "Recruiters always make the final call."
Mercor's AI interviewer follows a similar pattern: it asks role-specific questions, generates a transcript of your responses and evaluates your performance for the listing. Our Mercor AI interview guide has the details.
What it means for you: the rubric is usually written from the job posting, so the posting is your best preview of what will be scored. And because the AI can follow up, anything you mention can become the next question. Mention only what you can explain in more depth.
Set-up 2: AI scores a recorded one-way video
In a one-way video interview you see a question, get some preparation time and record your answer with no one on the other end (see what a one-way video interview is). Some employers add AI scoring on top. HireVue's 2025 Explainability Statement, still published as of October 2026, is one of the most detailed public descriptions of how that scoring works (HireVue Explainability Statement):
- Speech to text. A third-party speech-to-text system transcribes your spoken answer.
- Meaning. A language model reads the transcript for what your answer means, not just which words it contains.
- Scoring. A model trained on expert human ratings scores each answer against a behaviorally anchored rating scale for the competency, with five levels from novice to expert.
HireVue says this AI "relies only on what is said by the candidate" and does not assess facial expressions, body language, your background or tone of voice. If an answer has too little content to score, because it was very short, unclear or hit by a microphone problem, it is flagged for human review instead. Candidates who opt out of AI evaluation are rated by recruiters on the same scales. The report recruiters see lists the competencies assessed, how you scored on each and an overall result compared with the other candidates for that position.
Some places regulate this set-up directly. Illinois' Artificial Intelligence Video Interview Act, in effect since January 1, 2020, requires employers that use AI to analyze recorded video interviews for Illinois-based jobs to tell you beforehand, explain how the AI works and what general types of characteristics it uses, and get your consent. They must delete your videos within 30 days if you ask (820 ILCS 42). This is a summary, not legal advice. Check the law where you live, and see our overview of AI interview laws and candidate rights.
Set-up 3: an AI note-taker writes up a human interview
Here a person interviews you, but software records and transcribes the call and drafts the write-up. These "interview intelligence" tools change what the hiring team reads, even though a human gives the rating. As of October 2026:
- BrightHire says it structures notes by competency, maps them to the scorecard's rubric criteria and auto-drafts the scorecard into the employer's applicant tracking system, where the interviewer "reviews, adjusts if needed, and submits" it (BrightHire AI interview notes).
- Metaview says it can structure notes as Q&A or topic highlights, automatically fill scorecards, and sync notes and decisions to the applicant tracking system (Metaview interview notes).
The interviewer still decides your rating, but the notes shape what they remember and what everyone else reads later. An answer that never states your role or your result may simply not appear in the write-up. Our guides to AI note-takers in interviews and AI interview scorecards go deeper on this set-up.
Who makes the hiring decision
In most set-ups, a person does. HireVue's AI Interviewer page sums up its position as "AI Informs. Humans decide." Note-takers leave the rating to the interviewer by design. What varies is how much weight each employer gives the AI's output. Illinois' law even includes a reporting rule for employers that rely solely on AI analysis of video interviews to decide who gets an in-person interview, which shows lawmakers expected some employers to use AI scores as a gate, not just one input among several.
You are allowed to ask. A short, polite question to the recruiter works: "Is the AI score used to decide who moves forward, or does a person review every interview?" If the format puts you at a disadvantage, you can also ask about alternatives; see whether you can ask for a human interviewer instead.
What this means for how you answer
The three set-ups differ, but they reward the same habits, because all three start from your words.
| What the AI works from | What to do |
|---|---|
| A transcript of your words | Say your point in full sentences. Don't rely on gestures, slides or tone to carry it. |
| Criteria built from the job description | List the four to six skills the posting stresses and prepare one real story for each. |
| A rating scale per competency | Make the skill visible: what you did, why you did it and what changed. |
| Follow-up questions | Know the second layer of each story: the numbers, your decisions, what you would do differently. |
| A minimum amount of content | Give complete answers, usually a minute or two, unless the question only needs a fact. |
| Notes written from what you said | Name tools, metrics and your role out loud, and sum up your main point at the end. |
Don't stuff the posting's words into every sentence. HireVue says its model reads meaning regardless of the specific vocabulary used, and answers that sound read from a script or AI-written tend to hurt; many scorecard tools flag them. Use the employer's terms where they are true for you, in your own words. Our guide to what AI interview scoring looks for breaks the signals down one by one.
Before and after: one answer, all three set-ups
Question: "Tell me about a time you had to win over a skeptical stakeholder."
Before (example): "Yeah, so we had a project and the stakeholders weren't really on board, so we did a lot of meetings and eventually they came around and it went well."
After (example): "In my last role I led the rollout of a new returns process, and the warehouse manager opposed it because it added a step for his team. I asked for one week to run it on a single shift, timed the extra step myself and tracked repeat customer calls. The step took about 30 seconds per return, and repeat calls on that shift dropped. He agreed to roll it out to all three shifts, and I learned to bring data to the first conversation, not the third."
The second version works in every set-up. An AI interviewer can match it to a stakeholder-management criterion and has an obvious follow-up to ask. A one-way video model has enough content to score. A note-taker can quote the situation, your actions and the result word for word.
How to practice for all three set-ups
- Map the posting. Write down the skills the job description repeats. Those are the likely rubric rows.
- Draft one story per skill. The free STAR story builder gives each story a situation, task, action and result.
- Answer out loud and get scored. Our free AI interview assistant includes a Mock Interview mode that builds questions from your résumé and the job description, reads each one aloud and scores your spoken answer out of 10 with what to fix and a model answer, then gives you a report card.
- Re-answer your weakest question in your own words, using the fix note rather than copying the model answer.
- Try a second format with nothing to install. The free AI interview demo has 145 real interview questions across 10 tracks, graded by AI.
The free plan gives you 15 minutes every day with no card and no signup, which fits a short daily practice session. For a camera, microphone and room checklist, read how to pass an AI interview.
Hear how your answers score before an employer's AI does
Answer realistic questions out loud and get a score out of 10, with what to fix and a model answer for each one.
Try a free AI mock interviewFAQ
How do employers use AI to score interviews?
In three main ways: an AI interviewer that asks questions and rates each answer, AI scoring of a recorded one-way video, and an AI note-taker that drafts notes and a scorecard from a human interview. Most of these work from a transcript of what you said and compare it with criteria the employer set for the role.
Does AI make the final hiring decision?
Usually not. HireVue, for example, says recruiters always make the final call and treats AI scores as one input. Employers decide how much weight the scores carry, so ask the recruiter how AI results are used in their process.
Does AI interview scoring look at my face or voice?
It depends on the vendor. HireVue says its AI-scored video interviews rely only on what you say and do not assess facial expressions, body language or tone of voice. Other tools differ, so read the notice you are given or ask the recruiter.
Can I opt out of AI scoring?
Sometimes. HireVue says candidates who opt out of AI evaluation are rated by recruiters on the same scales, and Illinois requires consent before AI analyzes a recorded video interview for an Illinois-based job. Ask the recruiter what alternatives the employer offers.
What does the recruiter see after an AI-scored interview?
Typically a rating for each competency, an overall result, a transcript or recording, and sometimes a summary. When a note-taker is used, the hiring team sees structured notes and the scorecard the interviewer submitted.
How should I prepare for an AI-scored interview?
Read the job description, prepare one specific story for each skill it stresses, and practice answering out loud in complete sentences. A scored mock interview shows which answers need more detail before the real one.