Why Hiring Feels So Broken Right Now

Over the last few months, I spoke with more than 40 tech recruiters, CTOs, VPs of Engineering, and hiring leaders about how they experience the hiring market and the applications they get today.

Being personally frustrated about how bad hiring and getting hired currently is, I wanted to understand what wastes their time, what creates uncertainty, what has changed recently, and what they trust less than before. And also, why candidates are getting more and more frustrated.

Tl;dr: Most companies are not struggling because they have too few applicants. In many cases, they have more than enough. The bigger problem is that the application itself has become a weaker signal.

This also shows up in market data: Ashby found that applications per hire roughly tripled from 2021 to 2024 and stayed on average above 300 throughout 2025. (1)

And to clear a few myths before we go into the details: in my conversations, I did not see evidence that most recruiting teams are intentionally making this difficult for candidates. Most companies are not posting jobs just to collect data (they don't even use it that well), and ATSs are usually not some auto-rejection machine. The reality is more boring, and more frustrating: too much volume, too little signal, and too little trust.

Alright, that was a lot of preamble. Here are 5 reasons I found:

1. AI-written applications make it harder to know what is real

This was for sure the most common point.

CVs, cover letters, screening answers, and LinkedIn profiles are unbelievably polished with AI. I do not think this is automatically a bad thing. Candidates should be allowed to present themselves clearly, and many strong candidates (especially engineers) were never good at writing CVs or cover letters in the first place.

This is also not just anecdotal. Indeeds survey shows that 70% of job seekers use GenAI tools to research companies, draft cover letters, and prepare talking points.(2)

The problem is that polish can easily create the impression of substance, even when the underlying fit is weak. A mediocre candidate can appear much stronger than they are. A strong candidate can end up looking almost identical to everyone else. I’ve heard from more than one person that they received almost identical CVs from different people.

That drastically changes the first touchpoint. If recruiters or hiring managers can’t trust the initial signal of the application anymore, it shifts to mistrust and forces them to look for different signals. That is a very different hiring environment than it was before the wide adoption of AI.

I’ve heard one TA leader say “I’m now looking for imperfections in applications”... to determine whether it’s even worth to spend more time on an application. This is wild.

Business insider called it "'Grammatically correct and emotionally vacant'" (3). I like that.

2. Most inbound applicants are not qualified

Another theme that came up repeatedly was the amount of noise in inbound applications.

Several people told me that 70–90% of applicants are not meaningfully qualified for the role. Not slightly underqualified or maybe partially matching, but clearly not relevant. The mismatch was usually about seniority, location, work authorization, skills etc. On paper, a company may have hundreds of applicants. In practice, only a small percentage may be worth a serious look.

This leads to another problem: HR teams of larger corporations, that only look at the aggregated numbers, think that they “have enough” applications. But since the quality of applications dropped drastically, hiring managers are struggling more to fill roles.

That is what makes the current process so frustrating. Hiring managers and TA teams are now spending most of their time filtering out noise. They want to have meaningful conversations with interesting people and help them to find a new opportunity.

But the more noise there is, the easier it becomes to miss the good people inside the pile. That is bad for companies, but it is also bad for candidates, because strong applicants get buried in the same stack as everyone else.

As a personal anecdote: This is one of the worst feelings. After countless hours of screening 100s of applications, you’re still left with the feeling whether you missed the best candidate in all the noise.

3. Fake candidates are no longer just an edge case

This came up more often than I expected. Call me naive, but I just didn’t expect people to lie in their applications.

A lot of hiring teams mentioned fake or questionable candidates. They didn’t talk about candidates exaggerating their experience a bit, there were talking about situations where the person applying may not be the person who appears in the interview, where candidates seem to be located, where technical assessments are completed by someone else, or where profiles and work histories do not seem real.

I had such a situation myself, where a candidate claimed they’ve worked 10 years at Amazon and studied at an Ivy league university, which they clearly didn’t (that was easy to check). But it left me thinking how to avoid that situation.

For some companies, this is still rare. Or maybe they don’t notice, because they screen them out before. For others, especially in remote technical hiring, it has become something they actively think about and design their process around. Once fake candidates become a real concern, hiring managers naturally become more defensive. They add more checks and more skepticism, which often leads to more steps.

Codesignal reported that cheating and fraud attempts in proctored assessments more than doubled in 2025, rising from 16% in 2024 to 35%, driven by plagiarism, proxy test-taking, and unauthorized AI use. That would support your “not an edge case anymore” claim very well (4).

I’ve also now spoken to multiple companies that focus exclusively on building technology for candidate verification. They focus on just answering the question: "Is the candidate who they claim they are".

The uncomfortable part is that this affects everyone. Even honest candidates now go through processes shaped by the behaviour of bad actors. Soon, you’ll need to do additional tests, upload your ID somewhere, or maybe travel again to interview locations.

4. Candidates are often not prepared

Another recurring frustration was that many candidates show up with their hands in their pockets (figuratively).

Hiring teams described conversations with people who had not properly read the job description, did not understand the company, could not explain why the role made sense for them, or gave answers that felt generic and disconnected from the actual opportunity. Just some rehearsed ChatGPT output. At the same time, I think this problem needs some nuance. It is easy to blame candidates for being careless, but the system also encourages this behaviour.

When candidates feel that most applications disappear into a black hole, they apply to more roles. When applying becomes easier, they apply faster. When they do not expect a human response, they invest less in each individual application.

Over time, the whole market starts optimizing for volume.

Candidates apply to way more jobs with much less intent. Companies receive more applications with less relevance. Recruiters trust applications less. Candidates trust companies less... You see where this is going.

Everyone is reacting rationally to a system that no longer works very well.

Now, for the 5th point I could've chosen from many points that arguably would've been more consistent. People complained about using bad HR systems, bad processes, unclear job requirements etc. But I chose the more abstract next point. Although not fully consistent with the last four, it does summarise my experience talking to those TA professionals best.

5. Hiring teams are losing trust in the process

I’ve written about this before. Hiring is fundamentally about trust.

Recruiters do not fully trust the CV. Hiring managers do not fully trust the screening answers. Companies can’t trust technical assessments anymore. Candidates do not trust that a human will review them anymore, or that the role is real, or that the process will be fair. Once trust drops, everything becomes… slower? I also think that everyone loses when trust evaporates.

Greenhouse’s AI trust report says 91% of recruiters have found candidate deception, 34% spend up to half their week filtering spam and junk applications, 65% of hiring managers have caught applicants using AI deceptively, and 46% of U.S. job seekers say their trust in hiring decreased over the past year (5).

If companies are adding more steps. Candidates are applying more broadly. Recruiters add more filters. Hiring managers become more skeptical. Good candidates get frustrated, weak candidates continue to flood the process, and everyone feels like it’s the other sides fault.

That is the core problem I heard.

Most companies do not need more applicants. They need better signal. Something they can trust again. They need to understand who is actually qualified, who is genuinely interested, who understands the role, and who is worth spending time with.

AI has made applying easier, but it has made hiring much harder. Therefore everyone is suffering more.

That is where I think hiring needs to change. Not by adding more automation for the sake of automation, but by rebuilding the process around better signals: clearer role expectations, better candidate context, earlier qualification, more honest intent, and more transparency on both sides.

Oh... I didn’t expect anyone to read until the end.

This started as genuine curiosity, but after 40+ conversations about the hiring market, I think the patterns became very clear.

Maybe we can fix it together.

1. AI-written applications make it harder to know what is real

2. Most inbound applicants are not qualified

3. Fake candidates are no longer just an edge case

4. Candidates are often not prepared

5. Hiring teams are losing trust in the process

Sources