Hiring Ghosting Is Systemic. 140 Million Applications Confirm It.

Eric Thompson has a system. Every Monday morning, he opens his spreadsheet. Three columns: Company, Position, Date Applied. By Friday, he has added another 40 rows. He is 53, a cybersecurity professional with two decades of experience. Nine months ago, his startup laid him off. Since then, he has submitted 3,047 applications.

Five led to interviews. A hundred and sixty-seven generated any response at all.

That is a 5.5% response rate. For someone who spent twenty years keeping enterprise systems from being compromised by people who actively wanted to compromise them. The man could probably jailbreak the ATS that rejected him in less time than it took him to file the application. The irony is noted, at some length.

Thompson told reporters that half his applications generate no response at all, costing him roughly twenty hours of wasted work every week. Twenty hours researching companies, tailoring resumes, writing cover letters for jobs that were never going to be filled. He co-founded an advocacy group and started pushing for federal legislation. Federal legislation. Because companies were not sending rejection emails.

That is where we are.

What Thompson found, and what the data confirms, is that this is the system working as designed.

Where Technology Broke the Human Loop

There is a concept in economics called a phantom market: a market that looks active on paper but has no intention of clearing. It turns out the hiring version of this is now large enough to survey.

A Resume Builder survey of 1,000 hiring managers found that 40% of companies admitted to posting at least one job listing with no immediate hiring intent in the past year. They post them to benchmark compensation, to build pipelines for roles that might open up eventually, or simply because taking down a listing requires someone to remember to take it down. For whatever reason, candidates applying to these roles are performing unpaid labor for a company's internal research department. The company now has a very good sense of what they would pay for this person. The candidate has twenty hours less to show for their week.

The companies doing this are not, for the most part, consciously malicious. They poured billions into AI recruiting tools with the stated goal of freeing up HR teams for relationship-building and strategic work. What happened instead follows a pattern familiar to anyone who has watched an enterprise software rollout: the efficiency gains materialized, the headcount reductions followed, and the relationship-building never arrived to fill the gap. SHRM's 2025 Talent Trends report puts AI recruiting tool adoption at 51% of organizations, with broader industry surveys suggesting the figure runs as high as 60 to 70% depending on how you define AI screening. The candidate experience data suggests those tools are doing the screening. Nobody appears to be doing the communicating.

This tracks with a broader pattern in how enterprises have deployed AI. The tools handle the high-volume, low-judgment work reliably. The assumption that humans would naturally step into the resulting space and take on the higher-value work turned out to be optimistic. For context on how that pattern plays out across enterprise AI deployments more broadly, see our piece on the handoff problem in agentic AI systems.

The Arms Race in Both Directions

Here is the interesting part: both sides built this problem simultaneously, which is why neither side can solve it unilaterally.

On the candidate side: Greenhouse's 2024 platform data shows 38% of job seekers now mass-apply to roles using AI tools with minimal customisation per application. Application rates tripled between 2021 and 2024. By Q3 2024, the average recruiter was receiving 588 applications per open role, a 26% jump from the same quarter the previous year.

On the employer side: recruiting teams did not grow proportionally to handle that volume. They shrank. The AI tools were supposed to manage the gap. What they managed instead was a system that intakes applications and declines to acknowledge most of them.

The outcome: about 75% of applicants never hear back after applying, according to a Human Capital Institute survey reported by HR Dive. A May 2024 Resume Genius report found that 80% of hiring managers admitted to having ghosted candidates at some point. The most common reason given was that they were not certain they had found the right person and simply kept looking. Which is a remarkable thing to admit out loud, because the message that would resolve that situation takes about four seconds to send and costs precisely nothing.

For candidates who do make it past initial screens, the silence gets worse. 61% are ghosted after reaching the interview stage, up nine percentage points since early 2024, per Greenhouse's 2024 State of Job Hunting survey of 2,500 workers. Overall, a prospective applicant is now three times less likely to get hired than they were three years ago, according to Gem's 2025 Recruiting Benchmarks report covering more than 140 million applications.

Three times. In three years.

Hiring Ghosting by Group

The pain is not distributed evenly. Greenhouse data shows historically underrepresented candidates are ghosted at higher rates: 66% versus 59% for white candidates. An efficiency system specifically designed to remove human bias from screening has, in practice, automated the silence that follows screening and made the disparity in who receives it slightly larger. Worth noting.

At scale, hiring ghosting is an architectural failure. The pipeline was built to intake volume and filter output. Communicating with everyone who did not pass the filter was never in the spec. A form rejection email is not a courtesy. It is a basic output the system was never configured to produce. That is the thing that needs fixing, and it lives in process governance, not in cultural sensitivity training.

What Operators and HR Leaders Can Do Now

If you run a recruiting function, a staffing operation, or advise organisations on talent processes, the question is not "is this a problem?" at this point. The question is which part of it you are creating.

Three pressure points that cost nothing but attention:

Taking down job listings when a role is filled or paused is free. It requires someone to own the task. Assign it.

Configuring your existing ATS email automation to send status updates at each pipeline stage costs one afternoon and a conversation with your vendor. Most platforms have dormant communication templates that were never switched on during implementation. This is a fifteen-minute fix masquerading as a structural problem.

Setting a candidate response SLA, even something as simple as "we will contact you within 14 days if we wish to move forward," communicated upfront, eliminates most of the ambiguity problem at zero cost.

The hiring ghosting issue has grown large enough that federal legislation is being drafted around it. Getting ahead of that curve costs considerably less than responding to it. For a wider view of how remote and hybrid work reshaped organisational communication standards in the first place, see our analysis of the hybrid productivity gap.