As I noted in Low fire, low employment economy, the current hiring slowdown could make an already difficult job market even worse. Reliable job information is increasingly hard to find, while fraudulent listings and “ghost jobs” have become serious problems. Some states are even considering new laws to combat them.
Job seekers and employers alike say trust in the hiring process is breaking down. Candidates spend hours searching and applying for jobs, only to receive no response. For many applicants, the hiring process has become a black box.
Employers face challenges of their own. Some say they receive hundreds of applications that appear nearly identical. To sort through them quickly, they rely on applicant tracking systems (ATS) and artificial intelligence (AI) screening tools.
Chait says the job market is trapped in an AI doom loop. “We’re left with a tragic situation where both sides have a problem. They’re using AI to solve their problems, but the way they’re doing it is making the problem worse. So more use of AI means more use of AI, and it’s not in anyone’s interest. The more this happens, the worse the situation gets.”
Human review versus AI screening
I spoke with dozens of recruiters, human resources managers, and small-business owners who handle their own hiring. Some acknowledged using automated candidate-ranking tools, while others said they avoid them entirely. The choice did not appear to depend on company size or application volume. Instead, it seemed to reflect each organization’s philosophy and workplace culture.
Kim Jones, Toshiba’s vice president of human resources, says the company has a human review process for every application. She does not object to applicants using AI to polish their résumés or cover letters, but says doing so “doesn’t really help you get through the ATS.” Toshiba evaluates candidates based on several factors, including job requirements, salary expectations, and whether the person is eligible for rehire.
Jones has seen a more troubling use of AI during interviews. “When they hear a pause and maybe a typing sound, they come up with redundant answers,” she says, referring to candidates who appear to use AI in real time while responding to interview questions.
Doist is another company that relies more on human judgment than automated recruiting technology. The small, fully remote company hires workers from around the world, so applications can quickly overwhelm the recruiting team whenever a position becomes available.
Nadia Vatalidis, Doist’s human resources director, said her team tested whether ATS rankings could improve the hiring process. Using previously collected application materials and the original job description, they asked an AI system to shortlist candidates from the same applicant pool. They wanted to determine whether the people who had been interviewed—and ultimately hired—would appear among the top-ranked candidates.
In the two tests, the employees Doist hired did not make the AI-generated shortlists. There was some overlap with the candidates selected for interviews, but the eventual hires were missed. Vatalidis says those employees were performing well approximately six months after joining the company, highlighting the limitations of relying on automated screening alone.
How job seekers are using AI
When James Jacobsen began his job search five months ago, he spent nearly as much time searching and applying for jobs as he would have spent working a full-time position. As a design professional, he understands that AI cannot match his creativity, but he also recognizes that it can be useful for specific tasks. He used ChatGPT and Claude to improve his application materials and align them with job descriptions, but the strategy did not increase his results.
Jacobsen eventually shifted his focus from submitting more applications to managing his entire job search more efficiently. He used Claude to review job listings, analyze descriptions, and track opportunities from start to finish. He created a detailed scoring system based on job type, seniority, salary requirements, and whether a position was on-site, hybrid, or remote.
His AI assistant identified the strongest opportunities to pursue. When he decided not to apply for a position, he added a note explaining why. If the same job was reposted weeks later, the system could remind him of his earlier decision. This approach reduced the time he spent searching and helped him focus on higher-quality opportunities. Although he secured several part-time roles, he still had not received a full-time offer.
Source: www.wired.com