AI interviews can help hiring teams assess more candidates without adding more interview slots. But a faster process can still be a poor candidate experience if people do not know AI is involved, struggle to complete the interview, or never hear what happens next.
The answer is not to make the technology sound friendlier. It is to run a clearer process. For a better AI interview candidate experience, employers should do two things consistently: disclose AI use before the interview begins, and tell candidates when they can expect a status update. Then measure whether the process is working, rather than assuming that a completed interview is a good one.
What do candidates think about AI interviews?
Candidates are encountering AI interviews more often, but that does not mean they are comfortable with how employers use them. Greenhouse’s 2026 Candidate AI Interview Report, released on 1 May 2026, surveyed 2,950 active job seekers across the US, UK, Germany, Australia and Ireland. Greenhouse sells hiring software that includes AI interviewing, so its findings are vendor-sponsored research, not independent evidence. Unless otherwise stated, the figures below refer to the survey’s 1,200 US respondents.
The candidate view
Six numbers from 1,200 US job seekers
Source: Greenhouse 2026 Candidate AI Interview Report, released 1 May 2026. Vendor-sponsored survey; red marks the communication gaps.
These figures point to a practical problem. The interview format is only one part of the experience; communication before and after the interview also affects how candidates judge the employer. The report does not establish that disclosure alone prevents drop-off, but it gives employers a strong reason to make expectations clearer and measure what happens in their own hiring process.
1. Tell candidates when AI is involved
Disclose AI use before the candidate starts the interview. Explain what the technology does, what information it uses, and what happens after the session. Put that information where candidates will see it, not only in a long privacy notice. A candidate should not have to discover the format by clicking an interview link: make the disclosure part of the application journey, the interview invitation and the pre-interview instructions.
Minimum disclosure
What every candidate should know before starting
Tick off what your current invitation already covers.
Avoid vague statements such as “we use AI to improve hiring.” They do not tell candidates what will happen to them.
Use wording candidates can understand
Only include statements that accurately describe your setup. If the system evaluates specific skills, name them. If the employer has a defined retention policy, link to it. Do not promise a particular decision date unless the team can meet it. The wording also needs to match what candidates see on screen: if the invitation says one thing and the interview interface says another, the extra explanation creates more confusion, not less.
Copy-ready templates
Two messages to adapt
Switch between them. Highlighted text is yours to fill in.
About your interview
This stage of our hiring process uses an AI interviewer to ask questions and assess your responses against criteria related to the role. Your interview will be recorded. The resulting assessment will be used as part of our hiring process.
Please read the interview instructions and recording notice before you begin. You can complete the interview using the link provided. We will update you on your application status by [date or timeframe].
If you have questions about the process or recording, contact [team or contact details].
What happens after your interview
Once you complete the interview, our team will process your application and update you on its status by [date or timeframe]. If that timing changes, we will let you know.
You can contact [team or contact details] if you have a question about your application or experience a problem with the interview.
Put disclosure in the right places
One short explanation, three touchpoints
Keep it short and identical each time; link the fuller notice.
This is not about repeating a long disclaimer three times. Keep the explanation short and consistent, then provide a fuller notice where candidates can read it.
Fabric’s AI interview flow includes AI-use disclosure and recording consent in 35+ languages. Employers still need to make the surrounding communication clear: why the candidate received the invitation, what the interview is for, and when they should expect an update.
2. Explain what happens next, and when
Give candidates a clear timeframe for their next status update, then meet it. Candidates should not have to guess whether their interview was submitted successfully or when they will hear about their application. The Greenhouse finding that 51% of US candidates who completed an AI interview never heard back is specific to that survey, not a universal rate, but it highlights a basic process problem: an interview can finish while the candidate journey remains unclear.
Before launch
Decide these five things first
Each one needs a name or a number next to it.
Do not set an ambitious timeframe simply because the interview itself is automated. The rest of the hiring process still takes time. If the team cannot commit to a date, use a realistic timeframe and assign someone to monitor it. An automated acknowledgement is useful only if the process behind it can deliver what it says.
Treat delays as part of the candidate experience
A delayed decision does not always mean a broken process. Hiring managers may need more time, roles may change, and interview stages may be rescheduled. The avoidable problem is leaving candidates without an update. Create a simple trigger: if the promised update date passes, send a status message with a revised timeframe. Keep the language factual and do not imply that silence means rejection or progression.
The delay trigger
No candidate waits without a date
3. Make the interview accessible and practical
Check whether candidates can start, complete and resume the interview under realistic conditions. Accessibility is not only about formal accommodations; it also includes device access, connection reliability, language and clear instructions. Before selecting a platform, test the candidate journey on the devices and connections people are likely to use. Do not assume everyone has a laptop, a quiet home office or reliable broadband.
Pilot test
Test the journey before you invite at scale
Run each check on a real phone and a real connection.
A small pilot can reveal confusing instructions, browser issues and unexpected points of friction before they affect hundreds of applicants.
Be specific about language support
Do not describe a platform as multilingual and leave candidates to guess which languages are available. Publish the languages the product actually supports, and explain how candidates select one. Fabric supports 35+ languages, including English varieties (Indian, American, Australian, British and Singaporean), Hindi, Tagalog, Latin American Spanish, European Portuguese, German, Croatian and Japanese. The recruiter selects the interview language. Language support is not a promise that every language is evaluated in the same way, so make sure the language selected fits the role’s requirements.
Plan for interruptions and rescheduling
A candidate may lose connectivity, need to move an appointment or be unable to finish in one sitting. Your process should make the options clear before the interview begins. Fabric’s interview runs in a phone browser without an app, works on weak 4G connections, and resumes where the candidate left off if a connection drops. Rescheduling is allowed; retakes are at the hiring manager’s discretion, so define how requests will be handled consistently. For roles where a call suits candidates better, see AI phone screening.
Fabric
What happens when real life interrupts
4. Measure candidate experience, not just completion
Track whether candidates finish the interview, where they encounter friction, and how they rate the experience. Completion is useful, but on its own it does not show whether a process is clear, accessible or well communicated. Start with a baseline before changing the interview process, then compare results across similar roles, candidate groups and time periods. Where possible, compare the AI interview stage with the previous process while noting any differences in the candidate pool or hiring conditions.
Measure it
Six measures, and where to look when one slips
Set a baseline first. Define the denominator and time period before comparing; no single benchmark is a pass mark.
Avoid using a single industry benchmark as a pass-or-fail score. A completion rate can vary by role, candidate pool, interview length and the point at which the metric is measured.
Ask candidates short, useful questions
A short post-interview survey can help explain the numbers. Keep it focused on what the employer can improve, offer a simple rating scale and, if appropriate, one optional open-text question. Do not ask for feedback you will not analyse or act on.
Post-interview survey
Five questions worth asking
Try it as a candidate would. Add one optional open-text question at most.
Review the answers alongside operational data. A high completion rate with repeated complaints about unclear instructions is a reason to improve the experience, not to declare success. For a broader evaluation framework, see Fabric’s AI interviewer buyer’s guide. If you are still comparing types of platforms, the 2026 AI interview platform comparison explains the differences between the main categories.
Where Fabric fits
Fabric is built for teams that want to run AI interviews with access and communication designed into the process. Its candidate-facing capabilities include AI-use disclosure and recording consent in 35+ languages, phone-browser access without an app, support for weak 4G connections, and resuming after a dropped connection. Candidates can reschedule, while retakes remain at the hiring manager’s discretion. Recruiters can review every result, and scores cannot be changed. See Fabric’s AI interviewer.
Fabric
Two numbers to ask any vendor for
Fabric’s figures. When comparing vendors, ask how each figure is calculated, what sample it covers and when it was measured.
The employer still owns the candidate journey around the interview: explaining the process, setting realistic update timelines, deciding who sends status messages and monitoring whether candidates complete the process successfully. The right platform makes that work easier to manage. It cannot compensate for unclear communication or a hiring process that leaves candidates waiting without an update.
The next step
Candidate experience is shaped by what people know before an interview, how easily they can complete it, and whether the employer communicates afterwards. Start by making the two communication commitments explicit: disclose AI use before the interview, and set a realistic timeframe for the next status update. Then use candidate feedback and operational data to find what needs to improve.
Last updated: 26 October 2026.