AI hiring tools show racial bias, systemic rejection, and create a 330-day black hole. Here’s how to outsmart them.

Have you ever applied for a job only to get an auto-rejection letter within minutes or in the middle of the night? If you have, it probably gave you a sneaking suspicion that a human never actually reviewed your profile. Of course, this leads to frustration. But most job seekers aren’t going to demand to know if the hiring manager was reviewing applicants at 2 a.m. They’ll simply move on to the next “job opportunity.”

This turns into a cycle that repeats endlessly for many job seekers. Stanford knows why. Recently, the university released its findings from an extensive study looking at job applicants and AI hiring tools. The results were both astonishing and infuriating to those still looking for gainful employment.

The Stanford study wasn’t small. The university reports, “We follow 3.4 million people who submit 4 million job applications to 1,700 job postings across 150 employers and 11 industry sectors. Each job application was assessed by an AI hiring tool built by a single third-party vendor.”

Stanford found a concerning trend of racial bias when AI hiring tools were used. This bias wasn’t based on how the researchers felt; they utilized the formula provided by the Equal Employment Opportunity Commission (EEOC). The EEOC’s rule provides a clear approach to determine bias in hiring. Using the EEOC’s bias formula as a guide, Stanford found “that 26% of Black applicants and 15% of Asian applicants applied to positions where the AI system discriminated against their racial group.”

Racial bias is creating a gap

Those percentages may sound small, but they represent a large number. “To put this in perspective: If the AI had recommended Black and Asian candidates at the same rate as it recommended the most-favored group (typically white applicants), 40,000 more of their applications would have advanced to the next stage of hiring,” Stanford University Human-Centered Artificial Intelligence writes.

The authors of the study warn that when using averages, racial bias may seem nonexistent. This is because the algorithm does something seemingly nefarious. Of course, this is not personifying an AI algorithm, but a reminder that people write code. Sometimes, personal bias can sneak in without the intent of malice. In this case, Stanford says, “For example, imagine the AI tool frequently recommends Black applicants for warehouse jobs but rarely recommends them for finance jobs. If we were to average all the jobs together, those two patterns would cancel each other out and it would seem like there is no discrimination. The big-picture average hides the real discrimination happening job by job.”

The findings didn’t stop with race. They also looked at the large dependence on a shared hiring vendor by some companies. Stanford says, “We find that people who submit multiple applications to positions screened by the same algorithmic hiring vendor are more likely to be rejected from every position to which they apply than would be true if the companies made decisions statistically independently from one another.”

Essentially, if one company using this AI screening tool rejects an applicant, then every company that uses the same tool will also reject that same applicant. Each AI screening tool has a scoring process. That score stays with your information for up to 330 days, no matter how many jobs you apply for. If the new company you’re applying to uses the same AI hiring tool, your score continually gets recycled until the arbitrary time frame expires.

Finding ways around it

All of this may seem like a dire discovery, but there are ways around the dreaded black hole. A woman who goes by The Reframed Collective on Instagram shares that getting around the algorithmic monoculture will require some legwork from the applicant.

“The people who are getting in right now, they aren’t going through the front door. They are going around it,” she says. “You have to go and find that hiring manager or a network connection from that company who knows the hiring manager and can get you in front of them. And you have to do it as soon as you see the job posted. You have to ask a former colleague for an internal referral, and you have to get to a human before you ever submit a document. Because internal referrals are going to route a referral around an AI system,” she adds.

She also mentions the importance of using keywords directly from the job posting. Copying the entire job posting and pasting it into your resume is another trick, but be sure to make it “invisible” to bypass the first screening process. Based on comments, some people found success with this method while others said it did not work for them.

Beverly Dines, a Chief Empowerment Officer, shares that avoiding AI hiring tools altogether may yield more fruitful results. She reveals, “The algorithm cannot filter you out of a conversation that you’re already in.” The “CEO” then shares sites she has pre-vetted that will help you bypass the algorithm. “The first site is called Job Fair X. Job Fair X hosts over 2,000 virtual hiring events every year in both the US and Canada. Their hiring events cover multiple industries with actual folks in the room who make hiring decisions.”

Dines explains that after creating a profile with Job Fair X, you’ll be matched with companies that are hiring. The job seeker is the one who requests an interview, and, on the day of the event, the applicant is face-to-face with a human being, usually over video conferencing.

DiversityX is another company Dines recommends. She says the site hosts job fairs “for organizations that are focused on having a diverse talent pool.” Surprisingly, Eventbrite also made Dines’ list of places to get in front of actual humans who make hiring decisions. Using Eventbrite’s search bar, job seekers can pull up virtual and in-person job fairs in their area.

People shared their frustration in both of the women’s comment sections, with one person writing, “I’ve been actively applying to hundreds of jobs for 18 months, while still being employed. Where I would have gotten dozens of interviews five years ago, at minimal experience, I have now gotten a whopping 0 interviews. Craziest part? I’m lucky enough to have personal connections at one of the major medical device manufacturers that I’ve applied to twice, once without using the connection and once with using the connection.

The second time around I was told “you haven’t submitted your application yet, we need you to do that before we can set up a call”. I log in to the applicant portal, and sure enough, because it was the same role with same job ID as the one I’d applied to before, it would not let me re-submit an application, 8 months later. This shines the light on the fact that their ATS had auto-rejected me the first time around, despite being a perfectly-qualified candidate, and then HUNG ON TO THAT REJECTION SCORE UP TO 8 MONTHS LATER, still preventing my app from being reviewed by the internal recruiter. This needs to stop, NOW.”

Another writes, “This is pure insanity and HIGHLY illegal. As a former recruiter with decades of experience I believe lawsuits are the only way this gets reversed. Organizations need to be legally held accountable for their actions, because they don’t understand any other language.”

One person points out something employers may be overlooking. She writes, “If you’re an employer, this is keeping you from seeing employees you might actually want for a position.”