QUICK VERDICT
| Rating | 4/10 |
| Best For | High-volume, entry-level hiring where speed trumps candidate experience |
| Pricing | $5,000-$100,000+/year depending on company size and features |
| Strength | Processes thousands of applications that would otherwise overwhelm recruiters |
| Weakness | Research shows 85% bias toward white-associated names; candidates and employers both report the process is getting worse, not better |
Something has broken in the hiring process. Companies are using AI to screen the flood of applications. Job seekers are using AI to blast out more applications to get past the AI screens. Both sides are losing.
According to the Society for Human Resource Management, more than half of organizations now use AI to recruit workers. The Greenhouse 2025 AI in Hiring Report surveyed over 4,100 people and found that only 8% of job seekers believe AI algorithms make hiring fairer. Meanwhile, 91% of recruiters say they've spotted candidate deception. Daniel Chait, CEO of hiring platform Greenhouse, called it plainly: "Trust is at an all-time low for both job seekers and recruiters."
What I Looked At
I spent few days examining the AI hiring tool landscape, focusing on published research, court documents, regulatory filings, and survey data from multiple industry sources. I looked at the major players: Workday's screening tools, HireVue's video interview platform, and the broader ecosystem of applicant tracking systems that now touch nearly every application at large companies.
What I did not do: personally apply to jobs through these systems as a controlled test. I also did not have access to the back-end analytics that employers see. My analysis relies on public research, legal filings, and the considerable body of survey data from firms like Greenhouse, SHRM, and academic institutions. There's a lot of vendor marketing out there claiming incredible efficiency gains. I treated those claims with appropriate skepticism.
The Bias Problem Is Measurable
Vendors love to claim their tools reduce human bias. The research says otherwise.
A University of Washington study tested three large language models on resume screening with over 500 job listings and 120 names associated with different racial and gender groups. The results: the systems favored white-associated names 85% of the time versus Black-associated names only 9% of the time. Male-associated names were preferred 52% of the time versus female-associated names just 11% of the time. The intersectional findings were even more damaging: the systems never preferred Black male-associated names over white male-associated names in direct comparisons.
This isn't theoretical harm. The EEOC settled its first AI hiring discrimination case in August 2023 against iTutorGroup, a tutoring company whose software automatically rejected female applicants over 55 and male applicants over 60. The decree settling the suit provides $365,000 to be distributed to applicants who were automatically rejected due to age. One rejected applicant figured out the discrimination when they resubmitted an identical application with a different birthdate and suddenly got an interview.
The Workday lawsuit is the one to watch. Derek Mobley, an African American man over 40 with anxiety and depression, applied to more than 100 jobs through companies using Workday's screening platform. Rejected every time without an interview. In July 2024, a federal judge allowed the lawsuit to proceed, stating "Workday's software is not simply implementing in a rote way the criteria that employers set forth, but is instead participating in the decision-making process." The case was certified as a collective action in May 2025, potentially covering millions of applicants.
Video Interviews: The Loneliest Job Search
HireVue has conducted over 33 million "one-way" video interviews where candidates talk to a camera instead of a person. The AI analyzes facial expressions, word choice, and vocal tone to generate an "employability score."
In 2023, a Massachusetts resident filed a lawsuit against CVS Health alleging that the company utilized HireVue and Affectiva's software to assign applicants an "employability score," which was partly based on an applicant's "conscientiousness," "responsibility," and "innate sense of integrity and honor." CVS settled privately in July 2024.
The ACLU filed a complaint in March 2025 against Intuit and HireVue on behalf of a deaf Indigenous woman who worked at Intuit for years with positive performance reviews but was denied a promotion after HireVue's AI flagged her communication style negatively due to her deaf accent. The platform allegedly lacked proper captioning, and when she requested accommodation, it was denied.
HireVue did discontinue its facial recognition feature in 2021 after public backlash. The concerns were that candidates with accents or atypical speech patterns were being unfairly penalized. The company now relies on text-based analysis of interview transcripts, though critics argue the fundamental problem remains: you can't accurately predict job performance from a recorded monologue.
What Job Seekers Actually Experience
Only 8% of job seekers believe AI algorithms that screen initial applications make hiring fairer. Across all 1,200 U.S. job seekers polled by Greenhouse, almost half said their trust in hiring has decreased over the past year. Among Gen Z entry-level workers, that number rises to 62%.
The response has been predictable. Job seekers are gaming the system. Of the 1,200 U.S. job seekers Greenhouse surveyed, 41% admit to using prompt injections, hidden text designed to bypass AI filters, and of those who don't use this tactic, 52% say they are considering it.
LinkedIn applications spiked more than 45% over the past year, partly fueled by AI tools that let candidates blast out applications. But the flood creates its own problems. Among the influx of applications, most CVs were simple and nearly identical, as AI tools built them off the job descriptions rather than genuinely representing individual candidates. Greenhouse CEO Chait put it simply: "You end up basically not being able to tell anyone apart."
A majority (54%) of the US job seekers surveyed by Greenhouse said they've had an AI-led interview. Recent research from Dartmouth found that when job seekers use AI during the application process, they're actually less likely to be hired. The cover letters got longer and better-written, but companies stopped putting weight on them.
What Employers Actually Experience
Recruiters aren't thriving either. 34% spend up to half their week filtering spam and junk applications. The tools promised efficiency; they delivered volume.
Nearly half (49%) of U.S. job seekers apply to more positions just to get past automated filters. This creates a vicious cycle: more applications require more aggressive filtering, which makes job seekers feel they need to apply to even more positions.
The majority (68%) of hiring managers in the U.S. are more involved in hiring than they were last year. To verify authenticity, 39% of hiring managers in the U.S. are conducting more in-person interviews. So much for automation reducing workload.
Then there's the fraud problem. Over a third of job seekers (36%) in the U.S. have used AI to alter their appearance, voice, or background during video interviews. Deepfakes remain rare, but 18% of hiring managers report encountering them. The verification arms race consumes time that was supposed to be saved by automation.
Regulatory Landscape Is Shifting
New York City's Local Law 144 took effect in July 2023 and requires annual bias audits for automated employment decision tools. Companies must publish the results publicly and notify candidates when AI is being used to evaluate them. Violations carry penalties of up to $1,500 per violation.
EU AI Act obligations for general purpose AI began in August 2025, raising compliance expectations for employers and vendors deploying hiring technology. Several U.S. states are considering similar legislation.
The EEOC has been active. Beyond iTutorGroup, the commission has stated clearly that employers can't outsource liability to their vendors. As one employment attorney put it: "There's no defense saying that 'AI did it.' If AI did it, it's the same as the employer did it."
Pricing Reality
Enterprise AI hiring platforms typically cost $5,000 to $100,000+ annually depending on company size, features, and integration requirements. HireVue, Workday, Greenhouse, and similar platforms price based on number of hires, seats, or interview volume.
The promised ROI includes 25-50% reduction in time-to-hire and up to 30% reduction in cost-per-hire. But according to the 2025 SHRM Benchmarking Survey, average cost-per-hire and time-to-hire have both increased in the past three years, a period correlating with increased use of generative AI.
The math doesn't add up the way vendors claim. You might screen faster, but you're screening more applications than ever, dealing with more fraud, conducting more verification steps, and potentially building lawsuit liability.
vs. The Competition
The "competition" here isn't really between vendors. It's between AI-driven hiring and the traditional process that AI was supposed to improve.
Traditional hiring is slow and biased in its own ways. Recruiters spend 10 seconds on a resume. They favor candidates who remind them of themselves. The old system rejected plenty of qualified people.
But the new system doesn't appear to be better. A deaf woman with years of positive performance reviews gets rejected by an algorithm that doesn't understand her accent. A Black man with degrees from Morehouse and an honors graduate program applies to 100+ jobs and never gets an interview. An older applicant gets auto-rejected until they lie about their birthdate.
The vendor response is typically that their tools are "audited" and "tested for bias." But many of these audits are paid for by the vendors themselves. As one developer reviewing HireVue asked: "Would you tell a billion-dollar client that their system might be unfair to disabled or neurodivergent candidates if they're the ones keeping your lights on?"
What I Liked
- Resume parsing and basic matching saves genuine time on the most repetitive screening tasks
- Automated scheduling reduces email back-and-forth for interview coordination
- Bias audit requirements (where they exist) at least force some accountability that didn't exist before
- For truly high-volume, low-skill positions where speed matters most, the tools do process applications faster
What Needs Work
- Research shows systematic bias against Black, female, older, and disabled candidates that vendors haven't solved
- Video interview analysis remains pseudoscience according to multiple academic reviews
- The "efficiency gains" disappear when you account for increased application volume, fraud prevention, and manual verification
- Candidates report feeling dehumanized and distrustful of employers who use these tools
- Employers can't claim ignorance; they're liable for their vendors' discrimination
The Verdict
Daniel Chait warned that with AI infiltrating hiring from both sides, it has created a "doom loop" making everyone miserable. "Both sides are saying, 'This is impossible, it's not working, it's getting worse.'"
If you're an employer considering AI hiring tools: proceed carefully. The legal risk is real and growing. The Workday collective action could reshape vendor liability. NYC-style regulations are spreading. The EEOC is paying attention.
If you're a job seeker dealing with these systems: the frustration is legitimate. You're not imagining that your applications are disappearing into a void. 41% of candidates believe fewer than a quarter of their job applications were ever seen by a real person.
For high-volume, entry-level hiring where you need to process thousands of applicants and speed matters more than candidate experience, these tools have a narrow use case. For anything requiring judgment about complex qualifications, cultural fit, or potential? The technology isn't there. The evidence says human review, with all its flaws, still produces better outcomes for most roles.
The AI hiring industry is projected to exceed $3 billion. A lot of money is chasing the promise of automated recruitment. But only 26% of applicants trust AI to evaluate them fairly. That gap between vendor claims and user experience isn't closing. It's getting wider.
COMPARISON TABLE
| Feature | Workday | HireVue | Greenhouse |
|---|---|---|---|
| Primary Function | Full-stack HR/recruiting platform with AI screening | Video interview analysis and skills assessment | Hiring platform with ATS and recruiting workflows |
| AI Screening | Resume matching, candidate recommendations | Video/voice analysis, game-based assessments | AI-assisted screening (less aggressive than others) |
| Known Lawsuits | Class action proceeding (Mobley v. Workday) | ACLU complaint (Intuit case), CVS settlement | None major |
| Bias Audit | Vendor claims internal testing | Third-party audits (independence questioned) | Promotes "human-centric" approach |
| Enterprise Pricing | $100k+/year (enterprise license) | Per-interview pricing, enterprise contracts | $6k-$50k+/year depending on scale |
| Candidate Experience Rating | Low (see lawsuit specifics) | Low (one-way video criticism widespread) | Moderate (less automation) |
