I have spent years building and running hiring processes at multiple companies, scrapping unpaid take-home exercises, cutting five-round gauntlets down to three focused conversations, and treating candidates like professionals rather than resources to be evaluated and quietly discarded. I also mentor UX practitioners across this industry, and what I am watching happen to them in the current market is not bad luck. It is the predictable output of broken processes that were already failing before the economy got weird.
The market is genuinely hard right now. Candidates are spending months, sometimes over a year, in active search, navigating layoffs and hiring freezes and a political environment that has quietly rewritten the rules of who gets considered and how. That reality is real, and it is not the candidates’ fault. But a significant portion of the suffering in this market is not bad luck or bad timing. It is organizations running processes that were broken before any of this started, and those processes are getting worse.
Three fronts. All of them are operating simultaneously, and all of them are burning through excellent candidates while the people running these processes tell each other they just haven’t found the right fit yet.
Fear Is Running the Process, and Fear Just Rolled Back a Decade of Progress
I want to talk about two things that look different on the surface but are the same problem underneath. Fear-based hiring paralysis and the quiet dismantling of DEI infrastructure. They belong together because they share the same root cause. Nobody wants to be accountable for a decision.
Decision diffusion is what happens when nobody wants to own a hire. Not because the candidate is not qualified, but because making a decision means being accountable for the outcome, and accountability is uncomfortable when your company has spent three years laying people off and reorganizing reporting structures every six months. So instead of making a decision, committees add a round. They ask for more stakeholder interviews. They want to get alignment. They decide they need to revisit the role definition. And the candidate, who has now spent eight weeks in the process, gets a form email that says we have decided to pause the search.
This is cowardice dressed up as due diligence.
And it is the same cowardice that produced the DEI rollback. The political pressure on diversity programs over the last eighteen months has given a lot of organizations exactly the cover they were looking for. Structured interview panels get replaced with gut-feel conversations because nobody wants to defend the structure publicly. Evaluation criteria disappear. ERGs lose funding. And because none of it is announced as policy, nobody has to own it.
What structured criteria were doing was making bias legible and therefore accountable. A process built around gut reactions in an unstructured conversation tells you nothing you can actually use. It surfaces whatever the room already believed, and now nobody wrote it down, so nobody can be asked to defend it. A candidate from an underrepresented background makes it through the ATS, clears the recruiter screen, and then quietly stalls somewhere in the middle rounds. Not rejected. Not moved forward. Just gone. The feedback, if it ever comes, is a content-free “we decided to go in a different direction.”
I built a three-round hiring process at Paychex that reduced time to hire by nearly a month, scaled to seven simultaneous open reqs without breaking, and told us more about how candidates actually think and collaborate than the old five-round marathon. The reason it worked was simple: it was designed to make a decision. Three conversations, clear criteria, a cross-disciplinary panel, and then you either move someone forward or you tell them no.
The current market is running the opposite of this, and the candidates paying for it are the ones good enough to make it through the early rounds but not lucky enough to time out before the committee loses its nerve.
Credential Gatekeeping Is Class Bias With a Checkbox on It
I watched something happen recently to a UX practitioner I know in this market that I have not been able to stop thinking about.
Twelve years of real, measurable, portfolio-backed experience. And they lost a role after being in active conversation with a recruiter. Not because the work wasn’t there. Not because the interviews went badly. Because they have an Associate’s degree, not a Bachelor’s, and someone in the process decided that was a dealbreaker. In 2026.
Twelve years of demonstrated capability, dismissed because of a credential that has nothing to do with whether someone can lead a design system, run research, or build a team.
This is credential gatekeeping, and it is one of the most effective class-based filters in hiring because it looks like a standard. It looks like rigor. It is neither. It is an organization outsourcing its evaluation to a piece of paper because actually assessing whether someone can do the job requires effort, judgment, and the willingness to be wrong in a way you have to own.
UX specifically has one of the most legitimately nonlinear talent pools of any discipline. People come to this field from visual design, from anthropology, from library science, from product management, from theater. The practitioners who built this field largely did not have UX degrees because UX degrees barely existed when they were starting out. Some of the best people I have ever hired and managed came from paths that would not have survived a degree filter.
When you use a credential as a proxy for capability, you are not being rigorous. You are being lazy in a way that produces a paper trail that looks like a standard. And you are sending a message, loudly and clearly, to every experienced practitioner without a four-year degree that the door was never actually open.
AI Screening Is Doing the Filtering That Humans Are Too Scared to Do Themselves
Here is the part that should concern anyone who has ever thought seriously about what good hiring actually looks like.
Applicant tracking systems have always been blunt instruments. They were designed for volume management, not talent identification. Practitioners have known for years that keyword filtering eliminates qualified candidates, that the systems favor formatting conventions that have nothing to do with competency, and that the people who know how to game them get through while the people who write naturally do not.
What has changed is that most organizations are now using AI screening layers from the same small group of vendors. And researchers at Stanford just published data on exactly what that produces at scale.
The paper is called “Algorithmic Monocultures in Hiring.” It studied 3.4 million real applicants submitting 4 million applications, all screened by a single vendor’s algorithms across 156 employers. The researchers found that 25.87% of applications submitted by Black applicants and 14.74% of applications submitted by Asian applicants were directed to positions that adversely impacted those groups according to U.S. employment discrimination standards under Title VII. And then they found something worse: 4% of all applicants who applied to 10 positions were recommended for rejection from every single one, at a rate significantly higher than chance would predict.
They call this systemic rejection. The same algorithms, deployed across multiple employers from the same vendor, are flagging the same people as not recommended across every application they submit, regardless of role or company. Applicants would need to submit 25 applications to ensure at least one recommendation with 99.9% probability, compared to just 10 applications under independent decisions.
That is not a filter. That is a wall with a consent form stapled to it.
I built a skills library at Logility that encodes design standards into machine-readable instructions so AI-generated UI follows our system by default. I understand in practical terms how AI context shapes AI output. And what the Stanford research names is exactly what should concern every hiring manager who thinks they are running an objective process: when the same few vendors control how millions of candidates are evaluated before a human ever sees a resume, you do not have a hiring process. You have an algorithmic monoculture, and monocultures collapse in ways that are very difficult to trace back to any single decision anyone has to own.
The candidates being systematically eliminated by these tools are often exactly the candidates you most need. Nonlinear careers. Unusual combinations of experience. Practitioners who came to UX from somewhere unexpected and bring perspective your team does not have. The systems were never good at reading those candidates. AI has not fixed that. It has made the same mistake faster, at scale, and dressed it up as objectivity.
What This Costs
The people I mentor and talk to in this market are not amateurs who can’t find work. They are experienced, credentialed, portfolio-backed practitioners who are months or years deep in a search, navigating a system that was not built to surface the best candidates. It was built to reduce risk for the people running the process. Those are not the same thing, and pretending they are is costing organizations the hires they actually need.
When you run a process with no structured criteria, you get bias instead of information. When you design a process that cannot make a decision, you get attrition instead of hires. When you use a degree as a proxy for capability, you get pedigree instead of talent. When you filter candidates through algorithms that nobody in the organization fully understands or audits, you get elimination instead of evaluation.
And then the hiring manager tells someone on LinkedIn that the talent market is really tough right now.
It is. That part is true. It is also being made significantly worse by choices organizations are actively making about how to run these processes. Both things are true at the same time, and only one of them is in anyone’s control.
What Good Hiring Actually Requires
I have seen what it looks like when a process is designed to work. Structured evaluation criteria that every interviewer uses. A panel that includes the people who will work alongside the hire. No unpaid exercises, because asking someone to do a week of real work for free is extraction. Requirements that reflect what the role actually needs, not a credential checklist that filters for socioeconomic access. A timeline that reflects genuine urgency. And actual feedback to candidates who are not selected, because treating people like things is where the rot starts.
None of this is complicated. All of it requires organizations to make deliberate choices instead of defaulting to what is expedient, what feels safe, or what nobody will have to defend in a budget review.
The candidates who are getting eaten by this market are not the wrong candidates. They are the right candidates for organizations that do not have the courage to make decisions. And the organizations running these processes are not finding someone better. They are finding nobody, spending another six months requisitioning the role, and wondering why the same problems keep showing up on their quarterly engagement surveys.
The market is hard, and the processes are making it harder. Fix the processes, because we, as hiring managers, can at least do that much.
Amber Hansford is a Director of UX based in Atlanta who has spent fifteen years building design organizations and the hiring practices that sustain them. She mentors UX practitioners across the industry and writes about design leadership, DesignOps, and what it actually costs when organizations forget that candidates are people. The Stanford research referenced above is worth your time: algorithmichiring.github.io
