In the age of artificial intelligence, who gets hired, rejected, or fired increasingly comes down to algorithms. In a job market flooded with applications for every open position, AI tools promise speed and objectivity in screening candidates. But a wave of lawsuits reaching US courts, over age discrimination, lack of transparency in evaluations, and other issues, raises serious questions about how people find work today, or lose the chance at it, often without even knowing a machine is judging them.
The deeper problem is that most employers don’t build their own AI hiring systems but rely on the same third-party software. As a result, a negative evaluation can follow a candidate from company to company, a phenomenon researchers call “algorithmic exclusion.” In other words, hundreds of employers end up relying on the same “judgment” about a single candidate.
When a human rejects you in an interview, you can always try elsewhere. But an algorithmic system used by dozens of companies at once can “remember” an evaluation, and that negative signal can spread across the entire hiring pipeline, systematically shutting the same candidate out of many jobs. This raises a basic question: how much transparency and control should candidates have over machines deciding their professional futures?
The resume black hole
According to a Guardian report, Erin Kistler has applied to thousands of jobs over the past four years at companies like PayPal, Microsoft, and Netflix, only to watch her resume disappear into a black hole. As a product director with nearly 20 years of experience, she believes she was qualified for every role, yet was never called in for a single interview.
She is now suing Eightfold AI, the Silicon Valley company behind hiring software used by hundreds of employers, including ones she applied to, as part of a class action. The case was filed in January in a California court and is among the first to argue that automated candidate screening functions as a dossier on applicants, ranking them by predicted likelihood of success without giving them a chance to see or challenge the results.
Other workers are suing Meta over an internal AI system they say targeted them for layoffs because they had taken parental or medical leave. A recent lawsuit against IBM also claims AI tools discriminated against older workers. All three companies deny the allegations, according to the Guardian, with Eightfold and IBM calling them baseless.
A pretext for objectivity?
US companies are increasingly turning to AI for faster, more efficient workplace decisions, often under the banner of objectivity. But experts who study this kind of software say it can also introduce or worsen biases that follow candidates through every job search.
“There’s actually no law requiring notice or disclosure of the use of these AI hiring systems,” said Ifeoma Ajunwa, a professor at Emory University School of Law and founding director of the AI and the Future of Work program. “So companies don’t always tell employees when AI is being used to evaluate them,” she added.
Last year, 90% of employers used some form of automation in hiring, according to a World Economic Forum report. These tools range from basic filtering, like excluding candidates without a four-year degree or some other requirement, to using AI to assess a candidate’s skills or even conduct the initial phone screening.
In Eightfold’s case, the company describes itself as “the largest, self-refreshing source of talent data in the world.” It continuously updates an internal database with information from resumes, LinkedIn profiles, and social media profiles of more than a billion candidates who have applied for jobs through its platform.
Using this data, it applies AI to score candidates on a scale of 0 to 5, predicting how well they’ll perform in a given role. That score can determine which candidates get prioritized for interviews over others.
“A big part of the problem is that job applicants don’t know what’s in these dossiers,” said Rachel Dempsey, the lawyer representing Kistler. She argues that job seekers deserve the same transparency about the factors weighed in the hiring process, so they can either dispute inaccuracies or work to improve the things affecting their candidacy. Without any visibility, though, it’s impossible to know whether these algorithms are explicitly discriminating against them. “The idea of a ‘black box’ is deeply unsettling,” she said.
Bias reproduced
Almost all AI systems are trained to look for patterns, which can reinforce stereotypes or old biases. “As we’ve done more research on AI hiring systems, we’re finding that they actually tend to reproduce many of the same biases human managers have,” Ajunwa noted.
While human hiring managers can be unfair, say, by favoring a candidate who belonged to the same fraternity, “the scale of the bias is even bigger with AI,” said Xuechunzi Bai, an assistant professor at the University of Chicago who has researched hiring bias.
In a study Bai conducted using fictional demographic groups, AI models ended up “stereotyping” candidates based on irrelevant traits unrelated to their qualifications. The study found greater bias from AI compared to human hiring decisions examined last year, and newer, more advanced AI models produced hiring decisions with even more bias.
The consequences of AI bias for a candidate can extend well beyond a single job. According to Ajunwa, “you’ve essentially been placed on an algorithmic blacklist,” while Katie Creel, co-author of a recent study on the risks of “algorithmic exclusion” in hiring systems, stressed that people may end up shut out of jobs far more than they otherwise would be.
Some companies, like Incredible Health, insist that AI shouldn’t make the final call on how a candidate is scored or who advances to interviews, but should only assist human evaluators, an approach that has become a legal requirement in several US states.
In New York, a law that took effect in 2023 requires employers using automated hiring systems to conduct annual bias audits and notify candidates in advance about the use of the technology. The law, however, only applies to software that “substantially assists” or replaces decision-making, leaving a gap for cases where humans remain part of the process. Recent laws in Illinois and Colorado also ban employers from using AI tools that lead to unlawful discrimination.
