Interviewing
AI Interviews Are Common, and Mostly Vendors Have Counted Them
You did not get unlucky. In Greenhouse's 2026 report, 63% of 1,200 US job seekers said they had been interviewed by an AI; the vendor's November 2025 round put it at 54% [1][2]. Employer-side counts run lower and split hard by who is doing the hiring [3][4]. Every one of those numbers is a survey a vendor paid for, and as of April 2026 no federal statistic measured this [6]. Common is the honest answer. A precise percentage is not.
The takeThe exact percentage is the least useful thing to know here, and it is the number most write-ups lead with. The counts in circulation are mostly surveys run by companies that sell hiring software or job-search services. What is worth knowing instead is which format a particular employer runs, who reads the output, and at what stage a person gets involved. Those three are answerable by one message to the recruiter, and none of them turn on a national rate.
Where Olive fits
Open a role and see what the work shows
If an employer ever sends an Olive assessment, it 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 the employer reads, free, on every tier.
Assess your shortlistHow Common Is It, Really?
Common enough that meeting one is now the ordinary case for a US job seeker, at least on the only instrument that asks candidates directly. Greenhouse's 2026 Candidate AI Interview Report, released on 1 May 2026, says 63% of its 1,200 US respondents had been interviewed by an AI 1. Six months earlier the same company reported 54%, in the US findings of its November 2025 release 2.
Both are self-report. The respondent decides what counts as being interviewed by an AI, and neither release breaks the figure down between a chatbot screen, a recorded video answer graded by software, and a live AI interviewer 12.
The phrase is doing a lot of work. A scheduling bot that asks three knockout questions and a synthetic voice asking follow-ups can both land in the same box when a respondent answers yes. It is also a count of people who have ever met one, not a share of interviews run that way 1.
Neither release states a fielding window, a sampling method, or how the countries were weighted 12.
That matters more for the trend than for the level. The release calls it a rise of 13 percentage points in six months, which is not the gap between 63% and the earlier 54%, and either way it is one vendor comparing two of its own survey rounds 12. And the level is being reported by a company that sells hiring software to the employers running these interviews.
The exact number is soft. That the format is common is the part that holds.
Why Do Employer-Side Counts Run Lower?
Because the two sides are counting different things. A candidate counts encounters, and one employer running the format puts it in front of everyone who applies. Employer surveys count one answer per manager or company, however many applicants each one sees. A Resume Genius survey of 1,500 US hiring managers, fielded in June 2026, found AI-assisted video interviews reported by 40% of Gen Z hiring managers against 10% of Baby Boomers 3.
A ResumeBuilder.com panel of 948 US business leaders, run in October 2024, found about 23% saying their company uses AI to conduct interviews, with the phrase left for each respondent to interpret 4.
Both are panel surveys published by resume vendors, and neither is a probability sample of employers 34. The generational split in the first rests on uneven cells, the youngest of them small enough that the gap is a pattern rather than a rate 3.
The same round asked about disclosure. 43% of Gen Z managers said their company always discloses AI use, against 33% of Gen X and Boomers, and a separate 19% said their company uses no AI in interviews or candidate evaluation at all 3. Those two answers rest on different denominators and cannot be subtracted from each other.
Read the spread as a description of who you are applying to.
A large employer hiring at volume, with a young recruiting team, is a different proposition from a fifteen-person firm where a founder reads every application. The same job title in those two places can arrive with completely different first rounds. What employers are told about one of those first rounds is a different question: whether the async video round is worth keeping when the answers are scripted.
What No Survey Has Counted Yet
No official US statistic counted AI-conducted interviews as of April 2026, and the federal business survey of AI use cannot see them. The Census Bureau economists behind the 2026 AI supplement to the Business Trends and Outlook Survey wrote that their analysis gives no information on changes to hiring practices, and noted that questions on AI's effects on hiring are only planned for the next supplement 6.
That is a limit on one instrument, not a finding about the world. It does not say AI has no effect on hiring. It says this survey cannot see one.
What follows is narrow and useful. As of that April 2026 working paper, any figure offered as the official rate of AI interviewing did not come from the federal statistical system 6.
Adoption numbers for AI in recruiting are not adoption numbers for AI interviews either. LinkedIn, which sells hiring products, reports in its Future of Recruiting 2025 that 37% of organizations are actively integrating or experimenting with generative AI tools in recruiting, from a survey of 1,271 recruiting professionals across 23 countries in September 2024, each describing their own organization 5. That figure bundles pilots with production use, and it asks about recruiting generally, sourcing and outreach included, not about interviewing specifically. The same report gives no figure for how many organizations interview candidates with AI 5.
Three different questions keep getting answered with each other's numbers.
Worthless is still the wrong word for the vendor surveys. They are the most recent dated evidence on the question, weak as that evidence is.
Ask Which Format Is Coming, Not How Common It Is
The percentage changes nothing you do next, and the format changes most of it. At least four different things get called an AI interview, and what gets recorded, who reads it and how you prepare are different for each one. Ask the recruiter which one is scheduled and at what stage a person reviews it. That question is ordinary, and asking it is not a complaint.
Three things worth putting in one message back:
- Which format the first round uses, and roughly how long it takes
- Whether a person reviews the recording or the transcript, and at what stage
- What happens to the recording once the process closes
Declining is a separate decision, and whether you can turn down an AI interview and still be considered turns largely on whether the employer offers another format.
The time it takes is a real cost, and whether to spend it is yours to weigh. Treat it the way you would treat any other unpaid stage: the role, the employer, how far along the process already is, and how many other applications are open. A format being common is a fact about the market, not a reason on its own.
The invitation says more about the employer's hiring volume than about you.
Common questions
What percentage of companies use AI interviews?
No survey anyone can check gives a reliable figure. One employer-side number, about 23%, comes from a ResumeBuilder.com panel of 948 US business leaders in October 2024, where each respondent decided what using AI to conduct interviews meant 4. A Resume Genius survey of 1,500 US hiring managers in June 2026 reported AI-assisted video interviews used by 40% of Gen Z managers and 10% of Baby Boomers, which is a spread rather than a rate 3. Treat any single percentage as a survey a vendor paid for.
Is 63% the share of interviews that are run by AI?
No. It is the share of surveyed US job seekers who say they have ever been interviewed by an AI, from Greenhouse's 2026 report covering 1,200 US respondents 1. Someone who met one AI screen across fifty applications counts the same as someone who met ten. The figure describes exposure over a job search, not the composition of the interview market.
Does a government statistic exist for this?
Not as of April 2026. The Census Bureau economists who wrote up that year's AI supplement to the Business Trends and Outlook Survey stated that their analysis gives no information on changes to hiring practices, and that questions on AI's effects on hiring are only planned for the next supplement 6. Any figure offered as an official rate of AI interviewing is not coming from the federal statistical system. Check whether the next supplement has landed before relying on this.
Does 37% of organizations using AI in recruiting mean 37% run AI interviews?
No. LinkedIn's figure, from a hiring-products vendor, counts organizations its surveyed recruiters describe as actively integrating or experimenting with generative AI anywhere in recruiting, from a survey of 1,271 recruiting professionals across 23 countries in September 2024, and it bundles pilots with production use 5. Sourcing, outreach, scheduling and note-taking all sit inside it. The same report publishes no figure for how many organizations interview candidates with AI.
Am I more likely to meet one at a big employer?
Probably, though nobody has measured it directly for interviews. The one 2026 survey of hiring managers that splits the figure splits it by the manager's generation, from 40% among Gen Z managers to 10% among Baby Boomers on uneven cells, not by company size 3. Volume hiring is where an automated first round pays for itself, so a high-applicant role at a large company is the likeliest place to be sent one.
References
- 1. 63% of Job Seekers Have Faced an AI Interview. Most Haven't Had a Good One Yet greenhouse.com The 63% figure, its 1,200 US respondents, the 13-percentage-point rise, and the fact that the release breaks out no format detail behind 'interviewed by an AI'.
- 2. An AI Trust Crisis: 70% of Hiring Managers Trust AI to Make Faster and Better Hiring Decisions, Only 8% of Job Seekers Call it Fair greenhouse.com The earlier 54% figure, reported in the US findings of the November 2025 release, which is the round the 2026 release compares itself against.
- 3. 2026 Hiring Trends Report: AI, Ghost Jobs, & Getting Hired resumegenius.com The 1,500-manager June 2026 panel behind the 40% Gen Z against 10% Baby Boomer split on AI-assisted video interviews, and the disclosure figures quoted in the same section.
- 4. 7 in 10 Companies Will Use AI in the Hiring Process in 2025, Despite Most Saying It's Biased resumebuilder.com The most-circulated employer-side figure, about 23% of 948 US business leaders saying their company uses AI to conduct interviews, with the phrase left to the respondent.
- 5. The Future of Recruiting 2025 business.linkedin.com The 37% generative-AI-in-recruiting figure, its 1,271 respondents across 23 countries, and the absence of any interview-specific number in the same report.
- 6. The Microstructure of AI Diffusion: Evidence from Firms, Business Functions, and Worker Tasks (CES 26-25) www2.census.gov The authors' own statement that the analysis gives no information on changes to hiring practices, and that hiring questions are only planned for a future supplement, which is what supports the claim that no federal statistic covers this.
6 sources, numbered by first appearance. How Olive sources claims
General guidance for hiring teams. What works at one company and one volume may not transfer to yours.
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.