Job seekers are optimizing résumés for artificial intelligence screening systems that often don't exist, creating what one industry executive calls an "AI doom loop" that hurts both candidates and employers, according to an investigation published by Wired. The report reveals a fundamental breakdown in the hiring process: applicants invest heavily in tools designed to beat automated tracking systems, while many companies still rely entirely on human reviewers. The result is a market where trust has collapsed on both sides, applications have become indistinguishable, and neither party is getting what they need.
The investigation found that while some organizations use AI to automatically rank candidates—typically advancing only the top 10 to 20 percent of applicants for human review—others manage hiring personally from start to finish without any automated screening. Job seekers commonly pay $30 to $50 monthly for optimization services like Jobscan, which analyze résumés against job descriptions and suggest changes such as switching "percent" to the % symbol or limiting documents to one page. One data scientist interviewed received point deductions for using two pages and for including her middle initial inconsistently across documents. When one company tested its own AI ranking system against past hiring decisions, the employees they'd hired—people who were performing excellently after six months on the job—didn't appear in the AI-generated short lists in two separate trials.
The report documents how this mismatch fuels an escalating cycle. According to Daniel Chait, CEO of applicant tracking system provider Greenhouse, "We've got this tragic situation where each side has a problem." Job seekers believe AI gatekeepers control access to interviews, so they deploy their own AI to craft applications, polish cover letters, and even generate interview responses—behavior multiple human resources managers confirmed noticing through telltale pauses, typing sounds, and overly verbose answers. Employers, meanwhile, receive hundreds of nearly identical AI-written applications and respond by feeding them into automated ranking systems to find differentiation. Chait notes that each side uses AI to solve its own problem in ways that worsen the situation for everyone, with more AI use triggering even more AI use to no one's advantage.
The investigation concludes that the breakdown stems from a low-hiring economy where job postings are scarce, scams and ghost listings proliferate to the point that some states are drafting legislation, and applying has become what sources described as a black box with no feedback. One designer spent five months treating his job search as a full-time role, using AI to comb listings, analyze descriptions, score opportunities based on seniority and salary, and track rejections—essentially building a personal applicant tracking system in reverse—but received no offers despite revamping his portfolio and tailoring every submission. Industry leaders quoted in the report say this marks the first time both sides are simultaneously unhappy with the process. They recommend that candidates shift from high-volume applications to researching specific companies, submitting cover letters—which have become so rare they now stand out—and investing in networking, an underutilized strategy in today's market.
For hiring teams navigating competing pressures, the choice between speed and judgment has never carried higher stakes, while candidates face the uncomfortable reality that effort alone may not overcome systemic opacity. The message to job seekers, as Chait frames it: "It's not you; it's the system, and it stinks."