Policy
Ask for a Human Rater Before the Window Closes
You can ask for a person to score your interview instead of the AI. Send the request to the recruiter, in writing, before the interview window closes. HireVue's published statement says a candidate who opts out of AI evaluation completes the same assessment while recruiters rate the responses manually, though the employer configures what its consent screen offers. Another platform differs: HackerRank says a candidate who withholds consent cannot use that feature at all. Outside the EU, few laws require anyone to say yes.
The takeThe consent screen is the worst place to decide this. It arrives with a deadline, offers two buttons, and says little about what the second one costs. A recruiter's inbox is a better room for the same question, and it opens days earlier. A note that names what you want reviewed, and by whom, is ordinary process correspondence rather than a protest. Someone who never asks need not be afraid of the answer. The question can simply arrive five minutes before the recording starts.
Where Olive fits
Open a role and see what the work shows
An Olive assessment produces no composite score, and no decision about a candidate is made by a machine alone: a person writes or edits each of six findings, and each finding carries the evidence excerpt it rests on. The candidate is granted the same report as the employer, on every tier.
Assess your shortlistWhat Can You Actually Ask For?
Three different things, and naming the right one is most of the work. The narrow ask is a human rater on the same answers you were already going to record. The wider ask is a different assessment format altogether, which asks the employer to run a second process. The third is a human look after a decision has gone against you, which is a separate route with its own timing.
Start with the narrow one. You are not refusing the interview, not asking for more time, and not asking anyone to build a process that does not exist. You are asking who reads the recording.
Borrow the vocabulary from a platform that already publishes it. HackerRank's candidate AI notice names four things a candidate may request, and it points them at the employer rather than at the platform 3:
- an alternative selection process, or a reasonable accommodation
- human review of a decision made or substantially assisted by AI
- correction of inaccurate personal information used in that decision
- an explanation of how AI was used, after an adverse decision
Nothing in it obliges an employer to grant a single one.
Use the words anyway. A request phrased the way the platform itself phrases it is easier for a recruiter to recognize and pass along than one invented from scratch.
The platform supplies the form, and the employer decides.
Which Platforms Document an Opt-Out?
At least one does, in writing, so the vendor's name is the first thing to establish. HireVue's 2025 explainability statement says candidates see an AI consent statement before an AI-scored assessment and can opt in or out of AI evaluation. A candidate who opts out completes the same assessment, and recruiters manually review and rate the responses 1.
The same statement says the consent screen is required to tell the candidate that opting out will not exclude them from the hiring process 1.
Read that with the hedge attached. It is a vendor describing its own consent flow, and the screens are configurable by the employer deploying the platform, so what any one candidate sees may differ from the example in the document. The statement gives no figures on how manually rated candidates fare, or on how long a manual review takes. It covers HireVue and nothing else, and no other vendor's opt-out is described in it.
So the option may simply not be on your screen.
HackerRank publishes something different. Its knowledge-base FAQ says that when a customer switches on an automated tool, HackerRank secures the candidate's consent first, and a candidate who withholds consent cannot participate in the feature using that tool 2. That is not the same assessment rated by a person. It is that part of the process closed off.
So find out which platform is running the interview before you write anything.
The same HackerRank page draws a line worth knowing. It says the consent step is not intended to replace the ten-day advance notice and alternative-method request an employer must offer under the New York City law 2.
A vendor's consent box and a city's notice are two different obligations.
What Does a Human Rater Actually Change?
Less than most people hope, and the difference is worth understanding before you write to anyone about it. HireVue's statement says recruiters rating an opted-out candidate use the same behaviorally anchored rating scales 1. The instrument stays exactly where it was. The only thing that changes is the reader.
That distinction decides whether the ask is worth making. If the worry is that a model will read an accent, a pause, or a way of speaking that has nothing to do with the job, a human reader takes the model out of that step.
If the hope is a softer grader, nothing published promises one.
A human rater is also not the only human in the process. Whether a person ever watches an AI-scored interview is a separate question, and the usual answer is that someone reads your file later without watching the recording. Asking for manual rating changes who reads the answers, not whether anyone does.
What the swap costs in time goes unstated. A manual review waits on a person having time for it, and nothing in the statement says how long that takes, or how opted-out candidates fare next to AI-scored ones 1. No law fills that gap either.
The answers still have to be good.
Check Whether a Law Applies Where You Live
Three jurisdictions give the ask something to stand on, and none of them promises that a person will score your interview. New York City's Local Law 144 requires a candidate who lives in the city to be told at least ten business days before an automated employment decision tool is used. The same sentence says the notice must allow a request for an alternative selection process or accommodation 4.
The statute requires the employer to allow the request. It sets no deadline for an answer, does not say what an alternative has to look like, and reaches candidates who reside in New York City rather than everyone applying to an employer there.
In the EU, GDPR Article 22(3) covers a solely automated decision taken on contract or explicit-consent grounds. It requires at least three safeguards: human intervention on the controller's part, the chance to express your point of view, and the ability to contest the decision 5. Those attach only where the decision is based solely on automated processing and carries legal or similarly significant effects.
Colorado's replacement law, SB 26-189, signed May 14, 2026 and binding on no employer before January 1, 2027, works from the other end. A deployer must give a plain-language description of a covered tool's role within 30 days after it makes a consequential decision producing an adverse outcome, and the person may request meaningful human review and reconsideration 6. It is also not something you enforce yourself. The summary says the act creates no new private right of action, and the attorney general's rules on that disclosure are due January 1, 2027 6.
That is a route after a rejection, not before an interview.
Everywhere else in the United States, the ask rests on the employer's own policy.
The map is state by state, and it moves. What rights you have before an AI interview, by state and date sets out which jurisdictions require notice, consent or a waiver, and when each one started. Read your own row before you quote a statute at anyone.
The rights that come with applying to a European employer follow the employer's establishment rather than your passport.
Send the Request Before the Window Closes
A request made while the invitation is still open is a question about a process that is running. The same words sent after the deadline ask someone to reopen something they already closed, and that is a larger favor. Write to the recruiter as soon as the invitation mentions automated analysis, not after a week of thinking about it.
Keep it to three sentences. Name what you would like a person to review, say you are ready to complete the assessment either way, and ask what changes if they agree.
- Say: "The invitation mentions automated analysis of my answers. Would it be possible for a person to review my responses instead? I'm ready to record them either way."
- Not: a refusal to take the assessment, which is a different decision with different consequences
- Ask: "Does a manual review change the timeline on your side?"
- Not: a statute quoted at someone the statute does not cover, which reads as a threat and buys nothing
The timing question is the one people leave out. A manual review is somebody's afternoon, and the honest answer may be that it runs slower than the automated path. Hearing that before you choose is better than discovering it while the role closes.
Ask what it costs you, not only whether it is allowed.
If the answer is no, turning down the AI interview and staying in the running is the next question, and it is a genuinely different decision. If a decision has already come back, asking for a human review after an automated rejection runs on its own timing and its own tells.
Common questions
Will asking for a human rater get me dropped from the process?
Not according to the one platform that documents the option. HireVue's explainability statement says the consent screen is required to tell candidates that opting out of AI evaluation will not exclude them from the hiring process, and that an opted-out candidate completes the same assessment with recruiters rating the responses. That is one vendor describing its own flow, and consent screens are configurable by the employer deploying the platform. On other platforms, and wherever you ask the recruiter directly instead, the answer rests on the employer's policy.
Is there a law that makes an employer give me a human reviewer?
In the EU, GDPR Article 22(3) requires human intervention, a chance to state your point of view, and a route to contest, where a decision rests on a contract or your explicit consent, is based solely on automated processing, and has legal or similarly significant effects. New York City requires a ten-business-day notice that must allow a request for an alternative selection process, which is not a duty to grant one. Colorado's SB 26-189 adds a human-review request after an adverse decision, with the attorney general's rules due January 1, 2027. Elsewhere in the United States, no general right exists.
What if the platform says I cannot take the assessment without consenting?
Then the request moves to the employer. HackerRank's knowledge-base FAQ says a candidate who does not consent to an automated tool will not be able to participate in the feature using it, and the same answer says that consent step does not replace the ten-day notice and alternative-method request an employer must offer under the New York City law. Ask the recruiter for an alternative selection process rather than arguing with a consent screen, which has no third button. If the reason is a disability, say so and ask for it as an accommodation, which the employer decides. If you live in New York City, the notice must let you request either.
When should I send the request?
As soon as the invitation says your answers will be analyzed automatically, and before the recording window closes. A request made while the process is open is a question about how that process runs. The same request after the deadline asks someone to reopen a stage they have already finished, which is a bigger thing to grant and can land with a different person.
Does a human rater grade more gently?
Nothing published says so. HireVue's statement says recruiters rating an opted-out candidate use the same behaviorally anchored rating scales. The reader changes and the standard does not. A human rater takes the model out of the reading, which is the point if the worry is how a model handles an accent, a pause, or a manner of speaking. It is not a route to an easier assessment, and the statement says nothing about how opted-out candidates fare compared with AI-scored ones.
References
- 1. 2025 Explainability Statement hirevue.com Supports the opt-out path this article describes: a candidate who declines AI evaluation completes the same assessment, recruiters rate the responses manually against the same behaviorally anchored scales, and the consent screen must say that opting out will not exclude them.
- 2. FAQs: HackerRank Services and AI Laws support.hackerrank.com Supports the contrast with HireVue: a candidate who withholds consent cannot participate in the feature using the automated tool, and that consent step does not replace the ten-day notice and alternative-method request under the New York City law.
- 3. Candidate AI Notice hackerrank.com Supports the list of four things a candidate may request, and the point that the notice directs those requests to the employer rather than to the platform.
- 4. Int. No. 1894-A (enacted as Local Law 144 of 2021), a Local Law to amend the administrative code of the city of New York, in relation to automated employment decision tools dfs.ny.gov Supports the ten-business-day notice for a candidate residing in New York City, and the requirement that the notice allow a request for an alternative selection process or accommodation.
- 5. Regulation (EU) 2016/679 (General Data Protection Regulation), Official Journal L 119, 4.5.2016 publications.europa.eu Supports the three Article 22(3) safeguards named in this article: human intervention, expressing a point of view, and contesting the decision.
- 6. SB26-189 Automated Decision-Making Technology - Bill Summary (as enacted) leg.colorado.gov Supports the Colorado route this article describes as post-decision: a plain-language description within 30 days of an adverse outcome, a request for meaningful human review, and no new private right of action.
6 sources, numbered by first appearance. How Olive sources claims
General guidance, not legal advice. Hiring rules differ by state and country and change often; check anything here against your own counsel before you act on it.
Olive assesses how a person works with AI. It does not detect AI-written documents, and it never produces a score, a ranking, or a match percentage for a person. Candidates read the same report the employer reads.