Candidates walking out of AI interviews

Thirty-eight per cent of candidates have abandoned a hiring process because of an AI interview. That figure comes from Greenhouse’s 2026 candidate research, not from opinion. AI interviewing scaled faster than candidate tolerance for it. Employers now carry a measurable cost: applicants who stop replying. Here is what candidates reject, what they want instead, and where European rules push you.

What is an AI interview, and why do candidates walk away?

An AI interview is a screening stage where software asks the questions, records the answers, or scores them. Candidates leave when no person appears anywhere in that loop. Greenhouse found that 63% of job seekers have now faced one, up thirteen points in six months. Meanwhile 70% were never told upfront that AI would evaluate them.

The objection is rarely the technology itself. Instead, it is the silence around it. Roughly 21% only discovered AI involvement once the interview had started. Furthermore, 33% named pre-recorded video interviews, scored by AI with no human present, as their reason to quit. Surprise does more damage than automation, as the survey findings show.

How many candidates abandon an AI interview?

Greenhouse surveyed 2,950 candidates across the US, the UK, Germany, Australia and Ireland. Of those, 38% have already walked away from a process involving AI interviews. Another 12% say they would. Only 28% moved forward after completing one, and 51% received no feedback.

What candidates report	Share
Have faced an AI interview	63%
Walked away from a process because of it	38%
Rejected AI-scored video with no human present	33%
Never told upfront that AI would evaluate them	70%
Received no feedback afterwards	51%

Source: Greenhouse 2026 Candidate AI Interview Report, published May 2026.

What do candidates want instead of an AI interviewer?

Candidates are not asking employers to switch off every AI interviewer. They ask for control and clarity. Specifically, 46% want the option to request a human interview. Another 44% want disclosure upfront. Meanwhile 39% want an explanation of what the AI measures. A further 38% want human review before any decision.

Notably, none of those asks requires abandoning automation. Each is a process change, not a technology change. Therefore the fix costs far less than the drop-off it prevents.

What does an AI interview genuinely solve

AI interviewing earns its place in high-volume hiring. It removes scheduling friction, absorbs time-zone differences, and asks every applicant the same questions. For an employer screening hundreds of applications across five countries, that first-pass consistency is real. It also shortens response times, which candidates value.

Problems begin when the tool moves from sorting to deciding. Rejecting a person with no human involved is where trust breaks.

Where the handover from AI interviewer to human belongs

The handover belongs before any rejection, not after it. Use the AI interviewer for first-pass screening, scheduling and consistency. Then place a person at every decision point. Because a machine-only rejection is the moment candidates stop engaging, that single boundary protects most of your funnel.

Does the EU AI Act require human oversight of AI interviews?

It will, and the timing has shifted. The EU AI Act classifies recruitment and selection systems as high-risk under Annex III. Those duties, including human oversight under Article 14, were due in August 2026. EU legislators have since agreed to defer them to December 2027.

Two rules already apply, however. First, the Act prohibits emotion recognition in the workplace, and Commission guidance confirms that hiring sits inside that scope. Any AI interviewer inferring mood or sincerity from a face or voice is therefore already off-limits. Second, Article 22 of the GDPR limits decisions based solely on automated processing with significant effects. A rejection qualifies. Human review is therefore the defensible position today, and a formal requirement soon.

How European employers can protect candidate experience

Candidate experience decides whether people finish your process. Four moves cover most of the risk, and each maps to what candidates said they want.

• Disclose early. Name the AI interview in the advert and in the invitation email.
• Explain the measure. Tell candidates what the system scores and what it ignores.
• Offer the human route. Give every applicant a simple way to request a human interview.
• Keep the decision human. Have a person review AI output before any rejection, then send feedback.

These steps travel well across borders, which matters when you hire in several jurisdictions. Our guide to moving talent across borders covers the wider mobility picture. Our piece on talent acquisition and employer branding shows how process design shapes reputation.

Building a hiring process people actually finish

AI interviews cut administrative load, and that is worth keeping. Yet the data shows the limit clearly. Remove the human moment entirely and you lose candidates, feedback quality, and, before long, compliance ground. Octagon Professionals works at that intersection of people, systems and borders. We reduce concrete risks: undisclosed automated screening, solely automated rejections, and inconsistent practice across jurisdictions.

We do this because we want organisations to grow, with full transparency and nearly forty years of European experience. Just as importantly, you keep control. You choose the roles, the criteria, the salary and the offer. We remove the administrative burden, not the decision. To build a compliant, human hiring process across Europe, talk to Octagon Professionals.

Frequently asked questions about AI interviews

What is an AI interview?

An AI interview is a hiring stage run by software rather than a recruiter. The system asks set questions, records video or text answers, and often scores them automatically. Employers use it to screen large applicant volumes quickly and consistently, usually before any human conversation takes place.

Can I refuse an AI interview?

You can. Many employers offer a human alternative when asked, and 46% of candidates say they want that option. In the EU, GDPR also gives you the right to contest a decision made solely by automated means. Ask the recruiter directly and in writing.

Do employers have to tell you an AI interviewer is being used?

In the EU, transparency duties and data protection rules push strongly towards disclosure before the interview. Practice still lags: 70% of candidates were never told upfront. Ask the recruiter whether AI will evaluate your answers and whether a person reviews the result.

Are AI interviews legal in Europe?

Yes, with limits. AI interviewing is permitted, but emotion recognition in hiring is already prohibited under the EU AI Act. Recruitment systems also count as high-risk, with human oversight duties now scheduled for December 2027. Solely automated rejections risk breaching GDPR today.

How do you improve candidate experience when using AI interviews?

Disclose the AI upfront, explain what it measures, and offer a human interview on request. Then have a person review every automated evaluation before a decision. Finally, send feedback to everyone. Those four steps address what candidates named most often in 2026 research.