Key takeaways:
- Nearly half of U.S. candidates have either walked away from a hiring process that used AI or say they would, according to Greenhouse’s 2026 Candidate AI Interview Report.
- Thirty-four percent of candidates say an AI interview left them with a more negative view of the employer, meaning your clients are damaging their own employer brands one automated screening at a time.
- Staffing firms that build and communicate a human-led candidate experience have a competitive strength right now that most are underselling.
The numbers from hiring tech are easy to celebrate in a vendor pitch. Faster screening, lower cost-per-hire, more applications processed per recruiter hour. So it’s no surprise that AI use climbed to 43% of HR workflows in 2025, up from just 26% a year earlier, according to SHRM’s 2025 Talent Trends report.
But as adoption has scaled, so has candidate resistance.
Candidate withdrawal suggests a pipeline problem
Job seekers are struggling to stand out in an increasingly automated hiring environment and growing more frustrated by expanded AI use in hiring, according to Gartner’s 2026 talent acquisition research. The Greenhouse 2026 Candidate AI Interview Report, which surveyed 2,950 active job seekers across five countries, adds numbers to that frustration.
In the U.S., 63% of candidates have been interviewed by AI in the past 12 months, a 12-percentage-point jump in just six months. Thirty-eight percent have already withdrawn from a hiring process specifically because it included an AI interview. Another 12% say they would if required to complete one. That’s half the candidate pool either walking out or standing at the door.
Pre-recorded video interviews scored by AI with no live interviewer present was the top reason candidates decide to leave a hiring process (33%), followed by companies failing to disclose how AI would be used (27%), AI monitoring (26%), and fully AI-led interviews (26%). What those triggers share is not a rejection of technology itself. They’re about not knowing what’s happening, who’s evaluating them, or whether any human has actually looked at their application. Efficiency is a good thing, but not if it leaves candidates feeling invisible.
How does this impact your firm? When qualified candidates go through an automated screening process before a recruiter has spoken with them, some of those candidates drop off. The roles that are already hardest to fill get harder. Time-to-fill climbs. Your team works harder to backfill what the process already burned through.
AI interviews are putting clients’ reputations at risk
More than a third of U.S. candidates (34%) said an AI interview made their perception of the employer more negative, while 38% came away with a better impression. That net of +4 is the thinnest possible margin, and it’s likely overstated given what candidates do after a bad experience.
Fifty-one percent of U.S. candidates experienced no outcome after finishing an AI interview, but 38% were completely ghosted. More candidates were left with no answer at all than were moved forward in the process (28%). This suggests a breakdown in the employer-to-candidate relationship, happening at volume, on behalf of your clients.
A 2025 Gartner survey of 2,918 job candidates found only 26% trust AI to evaluate them fairly. Another 25% said they don’t trust an employer as much when they learn AI is part of the evaluation process.
When candidates leave a client’s process with a negative brand impression, that affects your ability to place talent with that client at scale. The strongest candidates on your roster have options. A client whose employer brand is suffering from poor AI practices is harder to represent well.
What candidates are asking for is already what your firm does
Most candidates aren’t entirely rejecting AI, nor are they asking for less of it. They just want better human oversight (22%) and more transparency (21%).
The top comfort measure is the option to request a human interview instead of an automated-only process. Fifty-nine percent of U.S. candidates prefer their initial conversation with a company to be a live interview with a human recruiter. Only 12% prefer an AI-conducted first touchpoint.
This makes your recruiters the core value proposition of your firm, and right now it’s rare enough to function as a real differentiator.
Being the human alternative means saying so out loud
It all starts with how you communicate the hiring process to candidates before they begin. Seventy percent of U.S. candidates who experienced an AI interview weren’t told upfront that AI would be evaluating them. Fifty-seven percent think there should be a legal requirement for companies to disclose AI use in candidate evaluation. Telling candidates upfront where AI does and doesn’t factor into your evaluation is a simple way to stand out in a market where opacity is the norm.
Next, where the human touchpoint sits in your process matters. A recruiter conversation after an AI screening round is not the same as a recruiter conversation as the first step. If your differentiator is human-led relationships, that needs to show up before candidates have already decided what kind of firm you are.
Finally, this is a client conversation, not just an internal one. Clients deploying AI-heavy hiring processes for roles they’ve asked you to fill are affecting the candidate experience of the people you’re trying to place. Sharing candidate sentiment data with clients, framed around pipeline health rather than their HR choices, is a conversation worth having. Gartner’s research recommends that recruiting leaders allow candidates to opt out of AI interviews specifically to build trust about fair treatment in the hiring process.
Staffing firms have spent years defending their value against automation. Now, half the candidate pool is walking away from AI-only processes, trusting the technology less with every unanswered application. Agencies that embrace a human-first process, upfront communication, and a client conversation about what AI is costing the pipeline won’t need to work hard to differentiate. The candidates will do it for them.
