AI Interviews: What to Expect
How AI and one-way interviews work, the notice and consent candidates are owed, how to prepare, and the red flags employers should avoid in AI tools.
By the HireRabbit.AI team · Published · Last updated

On this page
A first-round interview with software is now an ordinary part of applying for a job. For candidates, that raises practical questions: what is happening on the other side of the screen, what gets recorded, and what are you allowed to ask for?
Employers have their own questions: what they owe the people they invite, and which tools they can trust. This guide covers both sides in plain English.
How common AI interviews have become
More than half of US job seekers have already sat through one. Greenhouse's 2025 AI in Hiring Report, based on 4,136 job seekers, recruiters and hiring managers across the US, UK, Ireland and Germany, found that 54% of US job seekers had encountered an AI-led interview.
Employers are putting AI inside the interview itself. In an October 2024 Resume Builder survey of 948 business leaders, 81% of the companies using AI in interviews said they use it to ask interview questions, and 24% said AI conducts the entire interview.
The trust gap candidates bring
Adoption has run ahead of trust. Pew Research Center's survey of 11,004 US adults, published in April 2023, found that 71% oppose AI making a final hiring decision and 66% would not want to apply to an employer that uses AI to help make hiring decisions.
More recent surveys point the same way. Gartner found in July 2025 that only 26% of 2,918 job candidates trust AI will fairly evaluate them. Greenhouse found that 87% of US job seekers say it is important for employers to be transparent about their own AI use.
Side by side, those numbers show candidates asking for two things: to be told, and to have a person make the call. Neither costs an employer much.
How a real-time spoken AI interview works
The older format is the one-way video interview: you record answers to fixed questions within a time limit, and recruiters review the recordings later. The newer format feels much closer to a phone or video call. A typical session runs like this:
- The recruiter sends you a link, often with a window of a few days to complete it.
- A notice and consent screen explains that AI will conduct the interview, what is recorded and how the results are used.
- You check your camera and microphone in a short device test.
- The AI interviewer asks questions out loud, listens and can follow up, for example by asking for a specific example if your answer stayed general.
- The session produces a record, usually a video recording, a written transcript and a structured report for the recruiter.
- A recruiter or hiring manager reads the transcript and report and decides whether you move forward.
Step 6 matters most. In a well-run process a person still makes the decision.
Your rights and the notice you should get
This section is general information, not legal advice. Rules vary by location and change often, so check the current law where you live or hire.
In Illinois, the Artificial Intelligence Video Interview Act (820 ILCS 42) has applied since January 1, 2020 when an employer uses AI to analyze recorded video interviews for Illinois-based positions. The employer must tell you beforehand, explain how the AI works and what general types of characteristics it evaluates, and get your consent. If you ask, it must delete your videos, including copies, within 30 days.
New York City has enforced its law on automated employment decision tools since July 5, 2023. If an automated tool substantially assists in screening candidates for jobs in the city, the employer needs a bias audit within the previous year, a published summary and notice at least 10 business days before use, including how to request an alternative or an accommodation.
In the EU, the AI Act lists recruitment and candidate evaluation systems as high-risk, with those duties applying from December 2, 2027. Since February 2, 2025, it has also prohibited AI systems that infer what a person is feeling at work, including in hiring.
Even where no specific law applies, you can reasonably expect a clear statement that AI is involved, an explanation of what is recorded and who will see it, an accommodation or alternative format if you need one, and a person making the final decision.
How candidates can prepare
Most of the preparation is the same as for any interview, and students will find more in our job search guide for the AI era. The differences follow from one fact: a transcript of what you said is often the first thing a recruiter reads.
Set up your space. Find a quiet room, because background noise makes transcripts less accurate. Sit with the light in front of you, use a wired connection or sit close to your router, and close apps that use bandwidth.
Get your audio right. Audio matters more than video, because your words become the transcript. Use headphones with a microphone if you have them, do the device test properly, and speak at a normal pace.
Answer with specifics. An AI interviewer will often follow a vague answer with a request for an example, so give one up front. For a customer support lead role, "I stay calm and listen" draws a follow-up. "Last spring a client threatened to cancel a contract over a billing error; I called them the same afternoon, got finance to issue a corrected invoice within two days, and they renewed" gives the reviewer something to work with.
Use numbers where they are honest, such as team size or volume handled. And it is fine to pause, since most real-time tools wait for you.
Ask for accommodations early. If a disability, speech difference, hearing impairment or anything else means the format may not work for you, ask the recruiter before the interview for extra time, a different format, captions or a live interviewer.
Know what is recorded. Read the consent screen instead of clicking past it. It should tell you whether video or only audio is recorded, who can see it and how long it is kept.
What employers owe candidates
Get consent before the session starts. Say that AI conducts the interview, what it evaluates and how the results are used. In Illinois this is a legal requirement for AI-analyzed video interviews.
Explain what is recorded. Tell candidates whether you keep video, audio, a transcript and a report, who can access them and for how long, and meet Illinois' 30-day deadline for deletion requests.
Keep humans deciding. A recruiter or hiring manager should own the decision and be able to see how any score was produced, so they have grounds to disagree with it. If you are filling one operations manager role from 60 completed interviews, the reviewer who shortlists eight should be able to say why each made it.
Use integrity signals only as review flags. Greenhouse found that 36% of US job seekers had used AI to alter their appearance, voice or background during video interviews. Signals such as gaze leaving the screen can prompt a human to look closer, but they should never count against a candidate automatically. Keep them out of the score and let a reviewer judge them in context.
Offer alternatives and keep records. Have a real alternative, such as a live interview, for anyone the format does not suit. Retain the transcript, the report and the notice the candidate saw, so you can explain any decision later.
Red flags in AI interview tools
The most serious is a claim to read feelings or character from video. The science behind such claims is weak, they risk penalizing people with disabilities, accents or cultural differences, and in the EU inferring what people feel at work has been prohibited since February 2025. California's AB 1883 bans AI tools that infer how employees feel from January 1, 2027, though it is not yet clear whether it covers job applicants.
Other warning signs are a score with no breakdown, automatic rejection without a person reviewing the result, consent buried in a long policy, vague answers about retention, and monitoring signals that change the score instead of going to a reviewer. If you are choosing a tool, our guide on how to choose an AI interviewer lists the questions to ask vendors.
What a responsible setup looks like
At HireRabbit.AI, a recruiter sends a link, the candidate sees a consent screen before the session starts, and the interview runs as a real-time spoken conversation. The team receives the recording, the transcript and a scored report.
Eye, body and hand movement only create integrity flags for a person to review, and they never change the score. People make the decisions. The AI Interviewer is available through invite-only early access.
Candidates and employers can use the same three questions: were you told, do you know what was recorded, and is a person making the call? Employers can find the controls for the rest of the funnel in our playbook for hiring in the AI era.