"Recruitment Is Broken" — SHRM Argues the AI Arms Race Is Making Hiring Worse, Not Better
The News
SHRM published a provocative feature arguing that the escalating standoff between AI-enabled employers and AI-enabled candidates is eroding trust, fairness, and human judgment in hiring. The piece draws on SHRM's own 2025 Benchmarking Survey, which found that both cost-per-hire and time-to-hire have increased over the past three years — a period correlating with increased use of generative AI in talent acquisition.
The statistics back up the frustration: 40% to 80% of job applicants now use AI to write resumes, cover letters, or prepare for interviews. Meanwhile, 19% of organizations using automation or AI report that their tools have overlooked or screened out qualified applicants. And 78% of recruiting executives predict candidates will increasingly use AI to apply — setting up what SHRM calls an "AI arms race" that benefits neither side.
Quotes from analysts are pointed. Gartner's Jamie Kohn: "We are trying to make the process better, but at the same time, the technology is making the process worse." Lighthouse Research's Ben Eubanks: "There is no future where candidates using AI to beat AI creates a better outcome for hiring." And SHRM's own Nichol Bradford: "Recruiters can't go through thousands of applications. Job seekers are demoralized to never hear from a human."
Why It Matters
This is the counter-argument every TA leader needs to have ready — not because AI is wrong, but because the hype cycle has oversold what automation alone can do. The SHRM piece isn't anti-technology; it's anti-lazy-technology-deployment. The companies suffering from worsening metrics are the ones that deployed AI as a replacement for process improvement rather than an accelerator of it.
The nuance matters: the article acknowledges that high-volume frontline hiring is a legitimate exception where speed is the competitive advantage and candidates prefer AI screening for quick decisions. And AI can reduce human bias when properly designed — a human scanning 200 resumes sees a fraction of the candidates that a well-calibrated AI can surface. The lesson isn't "AI is bad" — it's "AI without process discipline, interviewer training, and candidate transparency makes everything worse."