Jodi Beggs' résumé got dinged for being two pages long. She used her middle initial on one document but not another, which also cost her points. If she changed “percent” to the % sign, her score went up.
Beggs, a data scientist and job seeker, got this feedback from an online system that compares a job description with an applicant's materials to assess how well they match up. The most well known of these tools is Jobscan, though Beggs didn't say which one she tested.
“If my job is to please the robots,” Beggs says, “then even if I don't like AI-generated content, maybe the machines do.”
Jobscan and other apps like it work on the premise that hiring committees use AI to automatically rate and rank candidates in their applicant tracking system (ATS). Only the top 10 to 20 percent of applicants are considered. The bottom 80 might never see the light of day. That's the belief, anyway. To get on the short list, candidates must tailor the formatting of their CVs and cover letters and pack them with the right keywords to match what the ATS wants to see.
The problem is, this premise isn’t always correct. Automated ranking absolutely happens in some organizations, but in others, hiring is still managed personally by humans from start to finish.
The Curse of the Applicant Tracking System
At its most basic, an ATS collects job applications, helps hiring committees review them, and tracks applicants through the interviewing and hiring process. Some systems offer more, from onboarding management to automated AI screening interviews.
Daniel Chait, CEO of an ATS company called Greenhouse, says there's a lot of folklore about these tools, particularly from the job seeker side. No two ATSs are the same. The technology changes fast, so an AI tool inside an ATS might work one way today and a different way tomorrow. Whether an ATS uses AI at all depends on the exact product and which features a team pays for and uses.
Regardless, as long as applicants believe AI is involved, they'll still try to game it using their own AI.
To put the fate of your job search into the hands of a tool that says it can optimize for an ATS is a shaky proposition. One recruiter proved how even with atrocious Jobscan scores, she still got a dozen interviews and an offer. Consider, too, that Jobscan costs $30 to $50 per month. The company isn't exactly incentivized to get you off the job market.
Being in a low-fire, low-hire economy only makes matters worse. Job postings are scarce. Scams and ghost jobs are enough of a problem that some states are looking to write new laws to combat them.
Both job seekers and employers say they're experiencing a breakdown of trust. Candidates sink so much time into their search and hear nothing. Applying to jobs is a black box.
Some employers, meanwhile, say they receive hundreds of applications that look nearly identical. So they feed them into the AI rating system of their ATS to try and get some quick differentiation.
Chait says we're in an AI doom loop. “We've got this tragic situation where each side has a problem. They're using AI to solve their own problem, but in ways that make the problem worse. And so more AI use begets more AI use, to no one's benefit. The more it's happening, the worse it gets.”
Comparing Human Vetting to AI
I spoke with dozens of recruiters and HR managers, as well as small business owners who do their own hiring. Some admitted they use automated candidate ranking, and some were adamant that they do not. The division has nothing to do with the size of the organization or even the number of applicants received. It seems driven by philosophy and culture.
Kim Jones is vice president of human resources at Toshiba. “We have humans review every application,” she says. She doesn't mind applicants using AI to polish up their materials, but “it's not really going to help with getting through the ATS.” Culling applicants comes down to job requirements, salary expectations, and a few other factors, like whether the person would be a rehire.
Where she has seen unwanted AI usage is in the interviewing stage. “You hear the pause, maybe hear the typing, then they come up with a verbose answer,” Jones says.
Another company that relies more on humans than technology for hiring is Doist. It's a small, fully remote company that hires internationally, which means when they have an opening, they're flush with applications.
Nadia Vatalidis, head of people at the company, says her team experimented to see if ATS rankings could help. They took roles that had already been filled and fed in the job descriptions as well as all the applicant materials that they had saved from the hiring process. They wanted to know “what would happen if AI had to short-list the same batch of candidates and whether the folks we interviewed would show up in those same categories,” Vatalidis says. “In two instances that we tested, the person we hired wasn't in the short list.” She says there was overlap in who got an interview, but their new hires—people who were working out great after about six months—didn't make the cut.
Using AI Differently
When James Jacobsen started his job search five months ago, the amount of time he sank into searching and applying to jobs was equivalent to a full-time role. As a design professional, he knows AI can't match him when it comes to creativity, but he experiments with it heavily and knows it's a powerful tool for certain tasks. He used Claude and ChatGPT to polish up his application materials and make sure they were in line with the job description, but it wasn't moving the needle.
Instead of focusing only on his application materials, he started using Claude to streamline and track his job search from top to bottom. He instructed it to comb job listings, analyze descriptions, and log them. He devised an elaborate scoring system based on the type of work, seniority, and his salary requirements, which varied for in-person, hybrid, and remote roles. His AI assistant highlighted the top jobs he should apply to. When he rejected one, he added a note, so if the same position got reposted weeks later, his AI could immediately remind him why he wasn't interested. He spent less time searching to find more high-quality positions. He got a few bites but no offers.
