We Didn’t Design Recruiting. We Inherited It.

Recruiting looks the way it does not because anyone designed it that way, but because it was inherited. Job adverts, resumes, interviews, and job descriptions were all borrowed from earlier eras and folded into hiring practice without much scrutiny of what they were actually for. That leaves deeply entrenched mental models about what recruiting should look like, so strong that even successful changes to behavior or systems tend to drift back to the old way of doing things.

A walk through recruiting history shows the pattern: persuasive letters functioning as early resumes, the first printed job advertisement, structured interviews built to test curiosity, and job descriptions created originally for wage benchmarking. More recently, applicant tracking systems reinvented the filing cabinet, job boards reproduced newspaper classifieds instead of rethinking them, and social recruiting became another channel for mass outreach rather than real connection. Each new technology got absorbed into old habits instead of prompting new questions.

Large language models risk the same fate if they are only used to speed up scheduling or rewrite job descriptions. Two examples point to a better path:

• A staffing company added AI screening immediately after application, giving candidates instant responses, surfacing transferable skills busy recruiters would otherwise miss, and freeing recruiters to build relationships instead of chasing paperwork across languages.
• A call center employer moved a soft-skills assessment to the very front of hiring, ahead of resume screening, while keeping recruiters responsible for the final interview and decision. Because the same leader also oversaw onboarding, recruiters could track how their choices played out in real performance and sharpen their judgment over time.

Both cases show technology and humans dividing the work so each does what it does best, rather than one replacing the other.

The risks are real. Automating a broken process just fails faster. Removing humans from decisions that are fundamentally personal, rather than transactional, damages both candidates and hiring quality. And a growing wave of AI tools on the candidate side, some resembling shortcuts or outright cheating, may end up driving as much disruption as anything on the employer side.

The real challenge is asking questions recruiting has never asked: whether a resume is needed, what an interview is for, why certain steps exist at all. Without that, new technology just automates old assumptions faster than before.