Job seekers are increasingly rewriting résumés to satisfy supposed automated hiring filters, even though employers use applicant-tracking systems in very different ways. Some organizations rank candidates with AI, while others rely on people throughout the process, leaving applicants to optimize for rules that may not exist.
Services such as Jobscan compare application materials with a job description and suggest formatting or keyword changes. Data scientist Jodi Beggs found one system penalized a two-page résumé, inconsistent use of her middle initial and even the word “percent” instead of a symbol. Such precise scores can create the impression that a universal machine gate controls access to interviews.
Recruiters say that belief feeds a cycle. Candidates use generative tools to produce more applications and tailor each one; employers receive hundreds of similar submissions and turn to automated ranking for differentiation. Greenhouse chief executive Daniel Chait calls it an AI doom loop in which each side’s attempted solution worsens the other’s problem.
The reporting also found no simple rule based on company size. Toshiba says humans review every application, while other employers openly use ranking systems. For applicants, the practical limit is that an optimization score cannot reveal how a particular employer configured its software. Clear evidence of skills and requirements still matters, and mass-producing machine-friendly prose may make an application less distinctive to the humans who ultimately evaluate it.