Why Some Candidates Reject Employers Using AI Interviews
Some applicants abandon AI interviews and avoid employers that require them. See what 2026 evidence proves, what it does not, and how employers can respond.
Yes—but the evidence supports personal avoidance, not a coordinated blacklist. Some candidates withdraw when an AI interview appears, refuse to apply again to employers that require one or keep their own notes on companies to avoid. No cited evidence shows a shared, industry-wide blacklist.
Choose what the employer has explained; the planner identifies what to ask before you proceed or withdraw.
AI Interview Decision Check
Use the employer’s actual disclosures, not assumptions about the software.
Ask for the missing details
The interview is still too unclear for an informed proceed-or-withdraw decision.
- Ask whether AI is used before you begin.
- Ask what information the system evaluates.
- Ask where human review or another interview route is available.
This planner organizes questions; it does not determine accommodation rights or whether an employer’s process is lawful.
Source basis: 2026 Greenhouse U.S. survey findings and U.S. EEOC disability guidance cited in the article. No industry-wide blacklist rate is available.
What the 2026 Evidence Shows
A Greenhouse survey found that 38% of U.S. respondents had withdrawn from a hiring process because it included an AI interview. Another 12% said they would withdraw if one were required. The survey covered 2,950 active job seekers in five countries, including 1,200 in the U.S. Unless otherwise stated, its published figures refer to that U.S. sample (Greenhouse, May 1, 2026).
That finding establishes candidate avoidance, but not the prevalence of employer blacklisting. The survey does not report the withdrawal rate per AI interview, how many employers each respondent rejected or whether candidates shared lists with other applicants. It is also a self-reported survey published by a hiring-platform company, not employer administrative data.
The clearest evidence of personal blacklisting comes from individual accounts. CNBC profiled three job seekers who described future AI interviews as a “hard no” or said they would avoid companies known to require them. One said he made a note of those employers and would not apply to them (CNBC, September 15, 2026).
Those accounts establish that the behavior exists. Three examples cannot establish how common permanent employer avoidance is across the labor market.
The defensible reading is:
- Some candidates are blacklisting employers for themselves.
- A substantial share of one U.S. survey sample reported abandoning at least one process because it included an AI interview.
- There is no cited evidence of a shared, systematic blacklist among candidates.
Candidates Are Rejecting Particular Processes, Not All AI
The Greenhouse results do not support the broader conclusion that candidates uniformly oppose AI in hiring. Only 19% of U.S. respondents wanted less AI. More commonly, respondents wanted the same amount with greater transparency or more AI with stronger human oversight.
The reported friction points were more specific:
| Issue | U.S. Survey Finding |
|---|---|
| AI use not clearly disclosed before the most recent AI interview | 70% of candidates who experienced AI evaluation |
| Prerecorded video scored by AI without a human interviewer prompted abandonment | 33% |
| Wanted an explanation of what the AI measured | 39% |
| Wanted the option to request a human interview | 46% |
| Completed an AI interview and never heard back | 38% |
| Completed one and were still waiting | 13% |
These percentages describe the Greenhouse survey’s U.S. respondents and the applicable subgroups, not all job seekers or every AI-interview system.
Employer reputation can move in either direction. In the same survey, 38% said a good AI-interview experience improved their view of the employer, while 34% said a poor experience left them with a more negative view (Greenhouse report summary). The implementation—not merely the presence of software—appears to shape the reaction.
“AI interview” can also refer to materially different processes. It may mean a voice bot asking fixed questions, an asynchronous recording evaluated by software or an interview in which AI only transcribes responses. Candidates should establish what the interview actually scores before treating every setup as equivalent.
One Bad Interview Can Become an Employer-Level Rejection
Candidates use the hiring process as evidence of how a company operates. An undisclosed bot, unanswered questions or an unexplained score can suggest that efficiency matters more than reciprocal evaluation. A candidate may then reject the employer rather than merely dislike the interview format.
The lack of a person can also prevent candidates from doing their own evaluation. A prerecorded interview does not let them ask about the manager, workload, team or role. If the company provides no recruiter contact or later feedback, the experience can feel entirely one-sided.
Technical and accessibility failures can deepen that impression. CNBC reported one candidate’s account of software repeatedly directing her to look straight ahead despite an eye condition. That is one person’s experience, not evidence about every AI tool, but it illustrates how a standardized format may fail to assess every candidate fairly.
In the U.S., the EEOC says employers should have a process for providing reasonable accommodations when using algorithmic decision tools. It also warns that such tools can screen out people with disabilities who could perform the job with or without a reasonable accommodation if safeguards are absent (EEOC guidance).
That guidance does not give every applicant a blanket right to reject any automated step and demand a human interview. Coverage and obligations depend on the employer, applicant, requested change and circumstances. State and local rules may add requirements.
A candidate facing a disability-related barrier can make a focused AI interview accommodation request rather than simply leaving the process.
What Candidates Should Establish Before Withdrawing
A recruiter should be able to explain what the employer means by “AI interview.” Candidates can ask whether the software presents questions, transcribes responses, monitors the session, produces scores or recommends who advances.
They can also ask what information is evaluated. The relevant distinction may be whether the system considers answer content alone or also analyzes delivery, audio, video or behavior.
Human review should be described precisely. “Human in the loop” does not explain whether a person watches the recording, reads a transcript, checks the score or can reverse a system-generated recommendation. Candidates can ask whether an AI output can prevent advancement without an independent review.
The available alternatives also matter. A candidate may ask whether the employer offers a live interview, written response or another format. An employer can decline a preference-based request, while a disability-related request may raise accommodation obligations depending on the circumstances.
Candidates should ask what happens to their data: who receives the recording, how long it is retained and whether it is used to train a model. The draft evidence does not provide a universal retention period or rule, so candidates need an answer tied to the specific employer, vendor and jurisdiction.
Refusing the automated step can still end an application. Disclosure duties, consent rules and accommodation rights vary, so candidates should distinguish a personal objection from a legally supported request. The separate guide to refusing an AI interview without automatically ending the process explains the possible routes.
Employers Can Reduce Candidate Avoidance
Employers do not necessarily have to abandon all interview automation. They do need to remove ambiguity that makes the process appear opaque, impersonal or inaccessible.
Disclosure should occur before a candidate schedules or begins the interview. It should identify what the system does, what it measures and how its output affects advancement. Calling a tool “AI-powered” without describing its role does not answer those questions.
Candidates also need a visible contact for technical questions and accommodation requests. Where an alternative format is legally required, it must be workable rather than nominal. Employers can consider offering an alternative more broadly even when it is not required.
Human review should be described in operational terms. Hiring teams should say who reviews the candidate’s response, what that person receives and whether the reviewer can override the system. Vague assurances about human involvement leave the consequential part of the process unexplained.
The assessment itself should measure job-relevant capabilities and be tested for accessibility and group disparities. Employers should also send an outcome after a candidate completes the interview; the Greenhouse findings show that silence after completion is part of the negative experience.
Finally, hiring teams can track withdrawals at the AI-interview stage by role and interview format. A rising withdrawal rate is a process signal. It should not be dismissed automatically as candidate resistance to technology.
Candidates are not uniformly blacklisting companies merely for using AI. Some are rejecting employers that surprise them with an opaque, impersonal or inaccessible evaluation. For hiring teams, the controllable risk is asking candidates to accept consequential automation without adequate notice, explanation, recourse or human accountability.
This article is informational, not legal or employment advice. AI-hiring requirements vary by jurisdiction and change frequently; confirm obligations for the locations and tools involved.