You were right

The buyers said it out loud

Since we started, job seekers have told us the same story in almost the same words: dozens of applications, weeks of silence, and a quiet suspicion that a machine said no before a person ever said hello. For a long time, the industry read that suspicion as sour grapes. This summer, the people who buy the machines said it themselves.

Paylocity surveyed more than 1,000 US HR and recruitment leaders for its State of Employee Recruitment report, covered by HR Dive on August 27. The numbers describe an industry all in: 91 percent actively use AI somewhere in their recruitment process, and the rest intend to within 18 months. 67 percent use it to screen resumes. AI resume screening is no longer an experiment at the margins. It is the default.

Then comes the sentence that matters. 26 percent of those same leaders say they are concerned about not seeing some candidates because of AI screening. Not candidates guessing from the outside. The people who run the filters, saying people go missing inside them. The same survey found 80 percent actively managing at least one issue with their AI tools, even as 85 percent say they trust AI to assess applicants regardless of background. Trust and worry, held at the same time.

Trust fell on both sides

Axios put it plainly in an August 21 headline: “The job application got faster. The job search got worse.” The piece cites Greenhouse’s 2026 AI in Hiring data: only 21 percent of recruiters were very confident their systems were not weeding out qualified candidates. On the other side of the table, 46 percent of candidates say their faith in the hiring process has decreased.

Volume explains some of it. Recruiters now process roughly 291 applications per hire, up from about 100 in early 2021 (Ashby data, reported in the same Axios piece). At that scale, something has to filter. And by Greenhouse’s numbers, few are certain what the filter leaves out.

More demand, flat hiring

iCIMS’s August 2026 Workforce Report, drawn from its own platform data across 691 million candidate profiles, found July job openings running 17 percent above the July 2025 baseline and applications up 6 percent, while hiring stayed flat year over year. That is the widest gap between employer demand and hiring activity iCIMS has recorded in 2026. More openings, more applications, no more hires. The distance between applying and being seen keeps growing.

Employ’s 2026 Job Seeker Nation survey found 34 percent of job seekers believed an AI had screened them out automatically. Set that beside Paylocity’s 26 percent of HR leaders worried about the same thing, and the two sides of the table are describing one problem.

Software owes people visibility

We believe a person should never vanish inside a process no one is checking. Software that judges people owes them visibility. That is not a feature request. It is the minimum.

It is why CoBlack works the candidate’s side of the same machinery. When Auto Match surfaces a role, it tells you in plain language why it fits, checked against your Career Capability Map. Capability, not keyword overlap. Openings come only from validated employer career pages and ATS feeds, so the job is real before you spend an evening on it. We automate the paperwork, never the judgment about you. The interview stays human.

And we are honest about the limits. We cannot change how any employer screens. What we can do is make sure you arrive relevant, and that nothing about you is ever invisible to you.

The suspicion was never sour grapes. It was an accurate reading of a system that has finally started describing itself.