For millions of applicants seeking jobs at hundreds of employers, the first hurdle is clearing an artificial intelligence system that screens their résumés and evaluates their suitability.
The process is similar, in some ways, to how credit agencies rank consumers by assigning them a numeric score based on their finances and borrowing history.
And now, a lawsuit filed by a group of job applicants claims that some A.I. employment screening tools should be subject to the same Fair Credit Reporting Act requirements as credit agencies. The lawsuit’s goal is to compel A.I. companies to disclose more information about what data they are gathering on applicants and how they are being ranked.
The target of the suit is a screening company, Eightfold AI, that sells its technology as a tool for employers to save time and money. Using sources like LinkedIn, Eightfold has created a data set that it says encompasses more than “1 million job titles, 1 million skills, and the profiles of more than 1 billion people working in every job, profession, industry, and geography.”
When candidates apply for a job, Eightfold’s software evaluates their skills and the employers’ needs, then scores the applicants on a scale of one to five.
Job seekers say the screening tool can become an algorithmic gatekeeper, blocking candidates from advancing to a human hiring manager and giving them no feedback on their scores or how the rating was generated. If the tool is making mistakes, candidates have no way to correct them.
“I think I deserve to know what’s being collected about me and shared with employers,” Erin Kistler, one of the plaintiffs in the lawsuit, said in an interview. “And they’re not giving me any feedback, so I can’t address the issues.”
Ms. Kistler has a degree in computer science and decades of experience in the technology industry. Out of the thousands of jobs she has sought in the past year, which she has meticulously tracked, only 0.3 percent of her applications have progressed to a follow-up or interview. Several of her applications were routed through Eightfold’s software system.
A representative for Eightfold, based in Santa Clara, Calif., did not respond to requests for comment.
The lawsuit, filed against Eightfold in Contra Costa County Superior Court in California, is an early attempt at what employers and their lawyers expect will be a wave of challenges to the use of A.I. in hiring.
David J. Walton, a Philadelphia lawyer who works with employers on A.I. issues and is not involved in the lawsuit, said companies could make a valid argument that these tools are different than how credit scoring works. The hiring software, he said, can be viewed as simply ranking candidates in the same way a human recruiter might sort applicants into tiers of desirable and less desirable candidates.
Still, Mr. Walton said that as companies push the boundaries of what A.I. tools can do, they’re often operating in legally gray areas — especially around data privacy and technology that may illegally discriminate against people even if not explicitly trained to do so.
“These tools are designed to be biased. I mean, they’re designed to find a certain type of person,” he said. “So they are designed be biased but they’re not designed to improperly be biased. And that’s a very fine line.”
Ms. Kistler’s lawsuit, which was filed by Towards Justice, a nonprofit Denver law firm, with help from former lawyers at the Consumer Financial Protection Bureau and the Equal Employment Opportunity Commission, is taking a relatively novel approach to challenging A.I. technology.
It is among the first cases to invoke credit reporting laws as a way to try to protect applicants against what some might refer to as “black box” employment decisions, where the applicant is kept in the dark about why they were disqualified.
Congress enacted the Fair Credit Reporting Act in 1970, not long after credit reporting agencies began using computer databases to compile their dossiers of personal information and turn them into numerical scores. To protect people against errors in those records, lawmakers required reporting agencies to disclose that information to consumers and allow them to dispute inaccuracies.
The law is broader than the credit reporting requirements for which it’s named. It defines a “consumer report” as any gathering of information on someone’s “personal characteristics” that is used to determine their eligibility for various financial services or “employment purposes.”
“There is no A.I. exemption to our laws,” said David Seligman, executive director of Towards Justice. “Far too often, the business model of these companies is to roll out these new technologies, to wrap them in fancy new language, and ultimately to just violate peoples’ rights.”
Mr. Seligman believes the law requires both Eightfold and the companies that use its technology to disclose to applicants what data is being gathered and provide them the opportunity to dispute and correct inaccuracies. The complaint, which is seeking class-action status, asks for unspecified financial damages and an order that Eightfold comply with state and federal consumer reporting laws.
Other lawsuits have taken aim at A.I. software systems for perceived violations of federal and state anti-discrimination laws. The most prominent is a 2023 lawsuit against Workday in the Federal District Court in San Francisco that says the company’s system, another popular one for screening job seekers, illegally discriminates against some people like older job seekers, those with disabilities and Black applicants.
Judge Rita F. Lin rejected Workday’s motion to dismiss the case, finding that the plaintiffs’ evidence — including a rejection notice one job seeker received at 1:50 a.m., less than an hour after submitting his application — “plausibly supports an inference that Workday’s algorithmic tools disproportionately reject applicants based on factors other than qualifications, such as a candidate’s race, age or disability.”
In May, she granted preliminary approval for the case to proceed as a collective action that could potentially include millions of rejected job applicants. A Workday spokesman said the lawsuit’s claims are false. “Workday’s A.I. recruiting tools are not trained to use — or even identify — protected characteristics like race, age, or disability,” the company said in a statement.
A 2024 guidance note from the Consumer Financial Protection Bureau opined that dossiers and scores created for hiring purposes were subject to the Fair Credit Reporting Act and that the vendors who created them legally qualified as consumer reporting agencies. Those notes serve as warnings for companies, telling them how regulators intend to enforce the laws they oversee.
Under President Trump, the bureau changed its stance. Russell T. Vought, the bureau’s acting director — who has spent his tenure trying to demolish and close the agency — rescinded the guidance memo in May.
Corporate litigation typically takes years, and the case against Eightfold is unlikely to move fast. But the underlying issues have also been brewing for years.
Jenny Yang, a former chair of the Equal Employment Opportunity Commission appointed during the Obama administration and one of the lawyers representing the plaintiffs in the lawsuit, said the commission began studying algorithmic hiring systems more than a decade ago.
“We realized they were fundamentally changing how people were hired. People were getting rejected in the middle of the night and nobody knew why,” she said.
Stacy Cowley is a Times business reporter who writes about a broad array of topics related to consumer finance, including student debt, the banking industry and small business.
The post Job Applicants Sue to Open ‘Black Box’ of A.I. Hiring Decisions appeared first on New York Times.
