Ribbon and Greenhouse report a rise in late-night AI job interviews
WIRED profiles a side effect of automated hiring: interview slots creeping into the middle of the night. Tim Millard, a communications and marketing professional in Atlanta, spent his job hunt interviewing for a media relations role that asked him to record himself answering questions on camera rather than talk to a person. He had 30 seconds to think and two minutes to answer each prompt, and did the recording at 9:30pm because the invitation left the timing open and he wanted it over with. He was rejected by automated email in March.
Arsham Ghahramani, who cofounded voice-AI recruitment company Ribbon, says he noticed the late-night pattern almost as soon as the product launched. Across more than 500 companies that use Ribbon, 24% of AI interviews happen between 10pm and 2am local time; among Ribbon's manufacturing clients, that share rises to 35%. Ghahramani frames the option as useful for parents, hourly workers and people tied up on a factory floor or in a kitchen during the day, who otherwise could not free up 30 minutes before 10pm. Hiring platform Greenhouse sees a smaller but still substantial version of the same shift: 15 to 20% of candidates interviewed by its voice agent now book a night slot. Ophir Samson, who became head of Greenhouse's voice AI division after the company acquired his startup in May, says he does not think AI should replace human interviews.
Candidates are largely unconvinced. A Greenhouse survey from May found that nearly two-thirds of job applicants have now been interviewed by an AI agent, up 13 percentage points over the prior six months. Thirty-eight percent of American candidates told Greenhouse they had withdrawn from a hiring process specifically to avoid an AI interview, and another 12% said they would drop out if one were required. A major source of the unease, per the article, is not knowing how the recorded footage is judged or whether a missed buzzword can sink an application outright.
How the scoring actually works varies by employer. Ghahramani says a welder might be asked about technical skills or certifications, while a sales candidate could be scored on an 'enthusiasm' rubric; AI then reviews the footage against that rubric and hands a recruiter a score plus the recording, though WIRED notes that in practice many of those recordings go unwatched by a human. Samson describes Greenhouse's version as a resume add-on meant to surface high-scoring candidates who would otherwise be buried in the pile of applications, framing the real choice not as human versus robot but as being ghosted versus getting a shot at a second-round human interview. 'It's a shitty time to be a candidate,' Samson says. 'Your resume goes into a black hole. You're competing with thousands of candidates that are all using AI to write their resumes. So, good candidates can't stand out.' A year after losing his job, Millard is still searching; the experience became the basis for a one-man show, After Careful Consideration, that debuted at the Atlanta Fringe Festival in May.
Key facts
- 24% of Ribbon's AI interviews across 500-plus client companies happen between 10pm and 2am local time; among its manufacturing clients that share is 35%.
- 15 to 20% of candidates interviewed by Greenhouse's voice agent now schedule a night-time slot.
- A May Greenhouse survey found nearly two-thirds of applicants have been interviewed by an AI agent, up 13 percentage points in six months.
- 38% of American candidates told Greenhouse they had withdrawn from a hiring process to avoid an AI interview, and 12% more said they would drop out if one were required.
- Tim Millard, who did a late-night automated video interview for a media relations job and was later rejected, turned the experience into a one-man show, After Careful Consideration.
Why it matters
Automated video and voice interviews are no longer a novelty step early in hiring; they are becoming common enough that vendors can measure a distinct behavioral shift, candidates moving their interviews into the night. That is a structural change in when and how people apply for work, not just a story about one awkward interview.
Who it affects
Job seekers with inflexible daytime schedules, cited by Ribbon's Ghahramani as parents, hourly workers and people currently employed on a factory floor or in a kitchen, who gain a slot they would not otherwise have. It also affects the broader pool of applicants who are simply skeptical of being screened by AI at all, a group the Greenhouse survey suggests is large.
How to use it
The article does not describe pricing or plans for Ribbon or Greenhouse. What it does describe is the mechanics candidates should expect: a rubric that varies by role (technical certifications for a welder, an 'enthusiasm' score for sales), a fixed thinking window before each question, a fixed recording window, and a score plus footage passed to a recruiter who may or may not watch it.
How solid is it
The interview-timing figures come directly from the two vendors whose products generate that data, Ribbon (24% and 35%, across 500-plus client companies) and Greenhouse (15 to 20%), so they describe those companies' own customer bases rather than the labor market as a whole. The candidate-sentiment numbers (two-thirds interviewed by AI, 38% and 12% opting out) come from a single Greenhouse survey conducted in May; the source does not give a sample size or say whether the vendor data is limited to the US.
Risks and caveats
The source itself flags the core risk: candidates get little visibility into how footage is scored or whether a missed phrase can end an application, and realistically much of the recorded footage is never reviewed by a human. Samson's framing, an AI pass as insurance against being ghosted rather than a replacement for human judgment, is the vendor's own characterization of the trade-off, not an independent assessment.
“It's a shitty time to be a candidate. Your resume goes into a black hole. You're competing with thousands of candidates that are all using AI to write their resumes. So, good candidates can't stand out.”
— Ophir Samson, head of Greenhouse's voice AI division