AI Interviewers Gain Traction in Hiring, Raising Fairness Concerns

Employers worldwide are increasingly using artificial intelligence-based interviewers to screen candidates in early stages of recruitment for full-time roles below director level. Companies such as Express (part of ManpowerGroup), Coinbase, and Zapier have adopted this technology to handle large volumes of applicants more efficiently. For example, Express uses a digital interviewer named Sophie who conducts 30-minute conversations and can ask follow-up questions, helping identify candidates who might be overlooked by traditional screening. One candidate interviewed by Sophie was eventually hired as a senior AI solutions architect.

Coinbase, which receives about 1.5 million applications annually, launched a digital interviewer called Milo in August 2025 for most non-director roles, successfully hiring over 240 employees initially screened by the bot. Zapier reports that AI tools allow it to screen up to five times more candidates than usual, with about one-third of those advancing to human recruiters not initially identified as suitable based on resumes alone.

A survey conducted in April among nearly 3,000 active job seekers, commissioned by recruitment software company Greenhouse, found that 63% of respondents in the US had undergone AI-led interviews in the past year, with lower but significant rates in Germany, Australia, the UK, and Ireland. However, 38% of job seekers reported quitting recruitment processes at least once due to AI interviews, and 12% said they would avoid applying if AI interviews were required.

The debate over AI fairness echoes earlier discussions about automated hiring tools. Proponents argue that algorithmic screening reduces human bias, while critics warn that AI scoring may mistakenly rely on speech patterns or facial expressions unrelated to job performance. As AI becomes a standard part of interviews, candidates are using similar technologies to prepare by practicing common questions and refining behavioral stories. Experts caution that AI should assist preparation rather than generate answers, as rehearsed responses can sound robotic and harm credibility. Candidates in sensitive fields like government, security, or intelligence are advised not to input classified information into AI tools during preparation.